专利摘要:
The present application discloses a system for effectively indexing advertising content using artificial intelligence to infer and generate thematic category groupings relevant to said content. These clusters are also matched to the collected customer profile data and are combined to deploy custom media solicitations through predictive analytics, addressing the problem of mass solicitations that are often inefficient and often poorly targeted. and therefore miss - their intended audience.
公开号:FR3016459A1
申请号:FR1550321
申请日:2015-01-15
公开日:2015-07-17
发明作者:Sebastien Plourde
申请人:INTEMA SOLUTIONS Inc；
IPC主号:

专利说明:

[0001] TECHNICAL FIELD [1] The present invention generally relates to various overlapping aspects of internet marketing, electronic commerce, and data analysis. Specifically, it's about post-click marketing and online lead generation, and includes custom media disciplines, marketing management, artificial intelligence, statistics, and analytics.
[0002] Background [2] The advent of the Internet, as well as the increased power of the underlying IT platforms on which it operates, have brought new opportunities for commercialization of products to target groups.  Online marketing has evolved in tandem with the evolution of the Internet itself, by combining different technologies and disciplines in order to come up with increasingly sophisticated methods by which one can better understand a target user or a database. to present to the said group advertising, information, or other content appropriate to it.  The proliferation of online shopping opportunities (e-tailing or online retail) for Internet users, in addition to the many new features that these experiences provide, has enabled online retailers and service providers to collect, analyze and, consequently, respond to information gleaned from both individual use patterns and collective trends.  Massive amounts of data collected on interest in a specific content or product can in turn be used by marketers to develop more appropriate product lines for a target audience or to target market segments , in order to increase the profitability of the company, in addition to promoting a beneficial competitiveness by being more sensitive to their respective markets, thus allowing a balance to be struck between innovation and a matching of the product offering to customer demand.  [3] As mentioned above, the computational foundations of the internet are particularly well suited to these tasks.  These technical basics have helped online advertisers develop marketing strategies of unimaginable granularity and precision before the widespread rise of the internet.  Indeed, a lot of information has been collected and analyzed, from which much has been learned.  As a result, various aspects of the advertising process have been automated and require little or no human intervention to function optimally.  [4] Despite this, some aspects of modern e-mail marketing strategies remain captive to a largely manual paradigm that requires the direct human participation of marketers and other administrative staff, who are forced to intervene in the assembly, management, rectification, optimization or when they are required to intervene in one or more critical facets - either for advertising campaigns or for their underlying marketing systems - and this, often fundamental way.  [5] A particularly regrettable limitation inherent in existing systems concerns the categorization of new advertising content within said systems.  This categorization may involve aspects where the advertisement (s) come to be categorized within a marketing system, and may require the extraction of attributes on elements such as content, including one or more aspects of the content, or the content of the content. -3 interior of which could prove appealing to customers in a hypothetical advertisement.  These elements typically require significant human intervention, the latter often taking the form of manual contributions by a human marketing director.  This manual human intervention is often a source of error and of a laborious nature, and the resulting system and commercial solicitations are limited and often deficient in their flexibility and scope.  [006] The needs of product creators or content creators, merchants and customers should and should be better served by the paradigms available to them, failing which these paradigms must be modified or rejected completely.  A system capable of sparing the content creator (or product marketer) the need to assign - often subjectively - categories or attributes to the product or content, or to spare those who fulfill these assign roles the laborious and error-prone tasks of determining in advance which user profiles from a large client bank (these profiles are often of dubious completeness and therefore reliability) could being interested in receiving specific content or solicitations for that content or the product could certainly improve the situation.  A solution to avoid unnecessary and / or poorly targeted commercial solicitations can save time, money, resources and avoid associated frustrations for recipients.  Similarly, a targeted commercial solicitation by which one or more potential buyers and sellers are gathered according to a known set of needs or interests of each of these buyers and the product offering each of these sellers could also bring a relief that both need greatly.  An improvement on one or both of these fronts can be achieved through more optimal automation and integration.  This automation can be achieved by linguistic analysis of the content from the product descriptions themselves, combined with an automated matching and delivery system better suited to one or more target audiences.  [007] This is an object of the invention described herein.  Summary [8] In an effort to spare marketing managers and content developers the onerous challenges of manually and sometimes subjectively identifying individual aspects of their articles - be they products or content or services - which may be suitable for their most appropriate clientele, the present invention having the modules and features as configured and summarized herein may be contemplated.  [9] Such configuration generally includes five key components (Fig.  1. ): 1.  An advertising content base (M01) receives and stores descriptions of products, article sheets, or so-called fundamental content that forms the basis of the marketing campaign for distribution to customers, also called users.  2.  A text analysis module (M03) receives the product descriptions from the advertising content base (M01) and performs a linguistic analysis of these descriptions.  The purpose of this analysis is to extract keywords as well as make inferences and generate category groupings for said content, these keywords and category groupings both forming important foundations for future advertising campaigns. .  The result of the analysis is a set of coefficients generated for the different attributes of the selected products, which attributes will be used in subsequent steps.  3.  A client attribute module (M13), which essentially stores the recipient profiles, including statistical information in the form of digital notation data and category data for each category generated at the text analysis module (M03), and is maintained and updated through a post-click history maintenance mechanism ideally collected over multiple marketing campaigns, said history to develop an accurate and continuously refined profile for each member within a set of clients.  Said profile is then used to predict or even anticipate which item sheets to use among a set of them in order to optimally solicit a set of customers with the modules listed later and further described in the present description, and this, on the the behavioral profile elucidated for each of said clients.  4.  A matching module (M09A) receives the attributes relating to the products of the advertising content base (M01), as well as the aforementioned product attribute weightings of the text analysis module (M03).  It also receives one or more client attribute data sets (M13), a cornerstone for determining whether one or more products having a given set of attributes can be suitably and predictively matched to a given set comprising one or more members of a target audience.  Unlike the various aspects of assessing product suitability for a target audience that was previously human-based, often on a static basis and for each campaign, the matching module (M09A) performs this assessment step by comparing dynamically the weighting of attributes assigned to a product with client attributes (including interest profiles), the latter having been collected by repeated use of the system by said clients.  These attribute data can be variously aggregated, either by the direct intervention of a human ad campaign administrator using an embodiment of the present invention, or dynamically through the machine learning capabilities. whereby an embodiment of the invention would have been configured.  In addition, said aggregation of data may be subject to various combinatorial configurations.  For example, the aggregation of profile attributes may result from the combination of several attributes, including behavioral characteristics collected from a single client.  Alternatively, such aggregation may take the form of a unique profile attribute collected from multiple clients.  More frequently, a specified amalgam of previous combinatorial arrangements is applied.  These arrangements, and particularly the information extracted therefrom, form the foundation of the aspect of predictive analysis encapsulated by embodiments of the present invention.  The implementation of said predictive analysis element is not the purpose of this module alone or any other single module described herein, but rather the cascading effect resulting from the interworking of several modules.  5.  The relative positioning of the content in a publication, whether it is an email message, newsletter, smart flyer, or other specified format - including but not limited to of a standard model among several possible arrangements - is specified, according to selection criteria used to feed and organize the arrangement of the elements in said publication, to the module - matching and layout template of selected products (M09B).  A key element of this module is the presence of a computer algorithm used to select, by ranking and matching (M09B1), the affinities linking the individual customers to one or more elements present in the advertising content base (M01).  The resultant of the aforementioned matching (M09B1) or matching process is the generation (M09B2) of a populated template based on said matching and matching operation, which resultant can be conditioned and transmitted accordingly, through one of several channels, including by email or the portal account of a user.  Therefore, the prediction-based matching enables the generation of personalized media, where the presence, placement, and disposition of the articles in said publication based on specified predictive analytics (or predictive analytics) criteria allow the resultant publication to achieve a higher marketing performance than the more naive mass-based commercial solicitation, due to the efforts made by the publication to anticipate needs and / or areas of interest on an individual basis for each customer or recipient to whom it is transmitted.  The present application describes an effective advertising content indexing system using artificial intelligence to deduce and generate thematic category groupings relevant to said content.  These category groupings are also matched to the collected customer profile data and are combined, for the purpose of deploying commercial media solicitations, through predictive analytics, addressing the problem of often inefficient mass solicitations. - 8 - who often target their desired clientele inappropriately - and therefore miss.  Although the present disclosure focuses on the marketing and solicitation of consumer goods - be it products, goods or services - it will be understood that the advertising content base (M01 ) should not be limited to such goods but may contain any other type of information, some of which may include non-limiting examples, such as newspaper articles, films, recipes, and images, any necessary modification to this effect made to the various modules described herein.  Brief Description of the Drawings [0012] The invention will be better understood by way of the following detailed description of the embodiments of the invention, which refers to the accompanying drawings, in which: FIG. block illustrating the main modules necessary for carrying out an advertising campaign by exploiting an embodiment of the present invention.  Figure 2 is a block diagram illustrating the key modules of an embodiment of the present invention, with particular emphasis on data routing and the inter-modular relationships underlying the data tracking capabilities. post-click of said embodiment for purposes of accumulating predictive analytics.  DETAILED DESCRIPTION [0015] In the following description, the use of the masculine or feminine gender is not intended to designate a particular genre, but refers to both men and women.  The arrival of each new technology provides a new means by which to chronicle the aspects of the world as perceived in a given time and place, or to convey visions or imaginary ideals of this world.  At the same time, and in virtually every society, advances in technology, commerce, and communication have promoted development, and more broadly, human evolution itself.  The evolutionary suite of primitive scribbles, cave paintings and artifacts from archaeological sites, ancient and modern architecture, existing mosaics, religious imagery, and even text and literature - bear witness to this evolution.  And through this evolution, factors of attraction and repulsion - whether in terms of technology, commercial dynamism, or even modes of communication - have sometimes been the subject of involuntary amalgams and other times voluntarily merged.  New messages, and occasionally new forms, have been the result of these combinations and mergers.  The combination and reuse forces are clearly visible in the historical progression of the means of communication, progression which includes rock paintings, engravings, sculptures, Renaissance art, printing press, radio and television.  A synthesis will allow to appreciate how all these forms have recently entered the digital age and have seen their underlying representation articulated - often retroactively - and combined involuntarily or deliberately amalgamated - by the binary logic.  Let us note that scholars have already observed that the communication medium through which a message travels when sent has the simultaneous effect of marking and shaping the relationship with this message and the way in which they are affected.  -10- [0018] It is therefore not surprising, perhaps, to note that with the arrival of each new technology, new ways have arisen whereby one can communicate on the trade, or more simply, with which one can advertise.  The fundamental purpose of marketing - and even a market in itself - is to promote the completion of transactions involving buyers and sellers.  The individual variants by which this objective has been sought and completed have certainly varied in space and time, but the underlying principle of linking supply and demand has remained essentially unchanged since prehistoric times.  The actual shape of the markets has evolved as the days and ten market districts of yesteryear have given way to centralized shopping malls and stores.  [0019] More broadly, the era of mass marketing has favored the growth of increasingly selective and demanding consumer basins, opening the way to increased competition between suppliers who, despite their sometimes global ambitions ( tacit or express) knew to compete for a set of customers, the latter is often limited.  Beyond its original meaning confined to a given place, the term "market" acquired an additional nuance, namely the pool of targeted people themselves.  Rather than spreading their message indiscriminately, the importance of courting (potential and existing) customers - 20 through better engagement in their own markets - has become clear to suppliers.  This has led to new ideas and practices to optimize the different aspects of commercial solicitation.  The process started by the print media broadcast won in a short time radio and television.  Put together, this trio represented a virtual monopoly of the "modern" advertising channels available to mass marketing professionals of the twentieth century.  Advertising has acquired a creative dimension, often combining market studies with non-empirical perceptions.  The trend towards deeper market research accelerated, as did the use of more sophisticated means of persuasion.  [0021] The internet has given new opportunities to marketers.  As was the case for the many means of communication that preceded it, a process of sophistication by which ways to better utilize (and abuse) this new technology is the subject of continuous exploration.  An example of this misuse of Internet technology for advertising purposes has been the proliferation of spam.  In fact, anti-spam technology, as well as the efforts of marketers to smarter their messages to more appropriate recipients, has ensued.  However, these attempts - as well as the related results - are far from perfect, as they still require human intervention, in whole or in part, to scrutinize various important aspects, either the advertising content itself or still the set of users to target (and often these two at the same time).  [0022] Careful human control of such fundamental aspects of a marketing campaign is deplorable for several reasons.  First, because contrary to the very nature of the internet, namely a very advanced communication medium whose structure and components are based on computational analysis, and on logic, the latter of a scale allowing deal with unprecedented amounts of data.  Second, it is highly unlikely that such human intervention, and the resulting decisions, will be in a manner reminiscent of their counterparts of the twentieth century, able to ensure that they are based on empirical evidence. , which can be a source of error, thus compromising the success of the advertising campaign.  Thus, even among the most advanced marketing systems, an important part of subjective human involvement in the meeting of buyers and sellers is likely to yield results that would in many ways be only marginally those of their predecessors.  The effects of poor marketing, particularly poorly targeted advertising - the solicitation of recipients who are not interested in a product or service - are fairly well understood.  From this point of view, the use of the Internet to disseminate advertising content indiscriminately and even carelessly to poorly chosen recipients is not only a flagrant and fundamental diversion of this powerful and new technology as well as a misunderstanding of its promise, but the application, in whole or in part, of anachronistic marketing strategies in this way often represents unnecessary efforts on the part of marketing specialists (especially in terms of time and money), and encourages flagrant and unthinking network bandwidth, energy, and time by all internet users.  This situation is unfortunately already too generalized and produces at best heterogeneous results.  And that is unacceptable.  [0024] The various embodiments of the online marketing system described herein provide a means by which to significantly reduce the inefficiency, laboriousness, and waste of resources resulting from the use of marketing systems. online contemporary.  This is achieved by reinventing the campaign design process to reduce human intervention, both in the critical phase of matching advertising content to potential recipients, and by inferring categorization techniques. intelligent means by which to index and retrieve said content.  These enhancements, affecting both the design and subsequent deployment of campaigns, play a major role, both in effective overall marketing and in the effective delivery of these campaigns in particular.  The data source upon which a large number of embodiments of the present invention depends is in the form of a pool of content, the essence of which is rationally related to the reasons for which the system is set up. as well as to achieve the overall goal of improving marketing effectiveness.  In many embodiments described herein, the content pool has the specific base designation of advertising content (M01), in particular because of the quality in which it serves as a collection for all commercial advertising content provided or available to the system.  In such embodiments, it must be appreciated that the precise meaning of the term "advertising content" is a deliberately flexible concept.  In one embodiment, such content may be introduced and stored in the database (M01) as discrete elements.  In another embodiment, such content may or may be aggregated and inseparable from particular products.  In addition, the concept of "product", as discussed herein, should, in a manner similar to that of the concept of "advertising content", also be considered in the broad sense and as having a non-restrictive scope.  Thus, both the type and nature of the articles being exchanged, sold, and / or negotiated through an embodiment of the present invention - particularly through biasing means discussed further herein - and contained in the database (M01), may encompass various forms.  A non-exhaustive enumeration of the object (singular or plural) of such potential transactions includes physical and concrete products and / or goods, in addition to more products of a more abstract, immaterial, and / or less reified nature, such as consumption. a news article, viewing a video, offering and buying a service, as well as any permutation of the foregoing.  Some examples of the product descriptions introduced in the advertising content database (M01) include, without limitation, elements such as textual information, image data, audio, and video recordings.  All of the above can be stored in raw or native format, or using any popular, proprietary format, or a format otherwise specified for that element.  The information represented by the content entered into the database (M01) by an administrator or other human operator over whom these privileges have been conferred may include, but not be limited to, a product description of the data sheets. containing information describing the product with varying granularity.  In some embodiments, the granularity of such information may provide, without limitation, one or more product designations or means by which to refer to it, in addition to other identifying information.  In various embodiments, this information may identify, without limitation, origin, price, manufacturer, designer, creator, supplier, resellers, dimensions, appearance, taste, and other attributes. said product which is considered appropriate, and relevant to the type of product described, as well as variants of each of the previous elements.  Depending on the embodiment, this description may incidentally or compulsorily contain photographic or video representations of the products, as well as variants thereof when available or deemed appropriate.  In addition, the content presenting a product or any element constituting a description of the product in the database (M01) may be present in one or more languages.  With appropriate modifications as appropriate, similar information may be provided when the subject of a commercial solicitation is a service.  In addition to the fact that no limitation is imposed, in the various embodiments of the present invention, as to the type and amount of underlying data formats that can be used to store the various Of the types of product description elements enumerated herein, a similar freedom characterizes the manner in which such discrete product description elements are organized, encapsulated, or mutually interconnected, whether by one or several human operators within said database (M01) itself.  In parallel, an unlimited number of otherwise disparate records existing in said database (M01) may, in another embodiment, be aggregated or merged by one or more human operators.  It will be necessary to appreciate this aggregation as being a notion largely fluid and whose main goal is to allow to regroup, in an external, explicit, and conceptual way, various elements in the database (M01).  The purpose of this grouping is to facilitate or further influence the operation of the text analysis module (M02) operation, described in greater detail herein, in its cybernetic extraction task of product categorizations. both meaningful and contextually relevant.  To cite a non-limiting example of such a human-specified aggregation, the mounting instructions for a piece of furniture may be explicitly associated with images of that article as well as a video to instruct viewers on the assembly of the piece of furniture.  As a corollary, a human operator may wish to dissociate a set of previously aggregated articles; this could be particularly desirable when a product is recalled or when a variant is discontinued and it is no longer practicable or even possible to maintain a previously valid aggregation.  As indicated above, a broad interpretation of the types of content that can be stored in the database (M01) is appropriate because the nature of one embodiment of the present invention should not be limited. artificially or even linked to a single family of advertising content.  It will be understood, however, that a broader appreciation of what is meant by advertising content in embodiments of the present invention makes it possible to deploy said embodiments in a variety of environments relatively easily.  Thus, in addition to or instead of the above-mentioned elements described above, a "product" need not necessarily be considered a tangible good, but in some embodiments it may equally be a service or a certain form of intellectual property, such as a written article or a video whose subject may be unrelated to advertising.  Moreover, while the present invention relates primarily to an advertising content base (M01) in the singular, it will be noted that such use encompasses only one conceptual entity and not a single material entity.  Thus, it should not be inferred that references to the advertising content base (M01) would in any way limit it to a single database or be interpreted in this way.  As a conceptual entity, the advertising content base (M01) may, in embodiments of the present invention, thus include one or more databases configured and arranged in one or more computers, interconnected by the intermediate of any network topology, operation based on any communication protocol, and having any organizational scope.  Similarly, the set of content stored in the database (M01) does not need to be static, but may change over time.  Therefore, the addition, modification and removal of content in said database (M01) may, in various embodiments of the present invention, be subject to an accessibility policy which would be limited to a or multiple human users, to whom one or more access types or profiles may be assigned whose privileges may vary in a variety of ways or confer consultation and / or modification capabilities.  As a non-limitative example of assigning profiles, the types of users authorized to access the database (M01) could be distributed among one or more content creators, whose work is scrutinized and controlled by a user. or multiple content editors, whose access accounts for them are created by one or more top administrators.  In another embodiment, the content submitted to the database (M01) may be modified manually or deleted, or, in another embodiment, be governed by an expiration date, after which said content is automatically deleted and deleted from the database (M01).
[0003] The parameters by which these deletion policies can be specified may be made available to human operators with advanced administrator privileges, and the deletion policies may further be adapted, to cite a non-limiting example, to accounts of users or specific types of content.  TEXT ANALYSIS MODULE [0033] The elements stored in the advertising content database (M01) are accessible by the text analysis module (M03), responsible for receiving the product descriptions of the content base. advertising (M01).  Therefore, the text analysis module (MO3) receives the product descriptions which are stored in the advertising content database (M01) and by one or more analytics techniques generates attribute values of weighted products extracted from said product descriptions.  The product attribute weightings -18- are used to determine whether or, in which measure, one or more products in the advertising content base (M01), can, as a result of their product attribute sets, be matched. appropriately with customer attributes (M13), the latter representing both known and statistically inferred interests of customer attributes of a specific customer base, these customer attributes being at least partly collected through processes predictive analytics (or predictive analytics).  The operation of this last module (M13) is addressed in the present disclosure; a discussion of the operation of the text analysis module (M03) will be initiated at the moment.  The work of the analysis (M03) text module consists largely of an extraction operation of natural language processing data and analytics, plus the task of performing said analysis on a or several sets of content which is textual in many cases, but which in some embodiments of the present invention is in no way limited thereby.  The database (M01) stores content whose organization especially with respect to the various pieces of advertising content for the same product - can be variously organized in a strong or weak organization.  The weighted product attribute values that are generated by the text analysis module (M03) represent the statistical statistical influence attributed to each of the descriptions in the database (M01) that have been made available to the module. text analysis (MO3) for keyword extraction purposes, and more specifically, for inferring and generating relevant thematic category groupings of said content (M01).  A key improvement offered by the present invention over prior online marketing systems is the comprehensive use of artificial intelligence to accurately perform two important tasks.  The first of these tasks is the extraction of sufficient information to derive one or more category descriptors from the product descriptions provided, while the second is to assign numerical statistical weighting to said category descriptors.  An iterative process by which numerical statistical weights are assigned to previously extracted category descriptors, including those derived for descriptions associated with other products, is also necessary in order to provision the remaining modules of the system described in detail soon, with the most appropriate product attribute weighting data.  The text analysis module (M03) achieves this by using the appropriate artificial intelligence, semantic analysis, and text mining processes that are robust and suitable for deploying a specific embodiment of the present invention.  Consistency in repeated cycles of this process, either in whole or in part, is also necessary to ensure the current nature of the resulting data; therefore, one skilled in the art will understand that in many embodiments, the atomicity properties of the databases must be fully taken into account.  An automatic learning algorithm suitably adapted to the types and formats of content stored in the advertising content database (M01) is deployed in the text analysis module (M03) for classifying said content.  A preliminary but essential aspect to consider is the ability of the algorithm to correctly interpret the different contents (M01) to which it will be applied.  In various embodiments of the present invention, the content may be transformed into a form suitable for analysis by said algorithm.  The appreciation of the adequacy in such contexts is a function of the ability of the algorithm to correctly interpret the different structures and content formats delivered to the text analysis module (M03) so as to generate groupings. relevant thematic categories, as described herein.  An in-depth analysis of the extracted information is performed on the product descriptions provided to the text analysis module (M03).  The precise type of analysis to be performed depends largely on the types of product descriptions provided potentially variable, but in many embodiments of the present invention, it is preferable to set up and deploy the corresponding components involved in the analysis. in a modular way.  These modules provide essential Natural Language Processing (NLP) capabilities and image processing to interpret the product descriptions provided to derive appropriate categorization terminology.  In various embodiments, several submodules, each dedicated to a specific task, interoperate to successfully generate said categorizations for a given set of product descriptions.  For this, the module must generally be equipped with various TLN capabilities, including a non-limiting enumeration that includes automatic summarization, speech analysis, machine translation, morphological segmentation, relationship extraction, and analysis. feelings.  In some embodiments of the present invention, relatively simple filtering and the deletion of unimportant words are typically performed.  Next, a language-specific radical search algorithm is applied, whereby related words are intelligently reduced to a common word root (eg "telephone", "telephoned", "telephone" and "telephones" all become " phone if necessary).  The weighting is performed on the sequence of words thus rooted, following which a comparison between the frequency and the repetition of words or other literal chains is performed.  The words or other chains of this kind are then sorted and pointed according to their particular relevance in the context of said article or description.  In various embodiments of the invention, each of these functionalities may furthermore be implemented and deployed as separate modules interconnected or integrated into a single module capable of learning and generating data. categorizations of the types described more fully herein.  Related activities may also be required and therefore deployed in the Text Analysis Module (M03) 10 since the product descriptions offered are specifically non-textual in nature (ex.  : images, audio or video), with the necessary adaptations, to text extraction and TLN features mentioned previously.  Such related activities may include optical character recognition, image analysis, speech segmentation, speech recognition, and other non-text related information extraction steps in order to correctly extract the information from the device. categorizing the descriptive advertising content in provided auditory and / or graphic form and assigning a statistical value for each category thus extracted from each of the various media types constituting the product description sources.  [0039] An important aspect in accomplishing the categorization and generation of the relevant thematic groupings described above is the continual ability to adapt a set of related analytics tasks to a set of potential product descriptions. scalable.  While many embodiments of the present invention utilize multiple forms of artificial intelligence rather than direct human intervention to generate categorization headings and weights, leaving such generation to Even the most advanced machine learning module of contemporary technology involves the risk of unforeseen, unexpected, and - more importantly - unwanted situations occurring both in the generation of categorization rubrics by the text analysis module. (M03) than in assigning weighted attributes to these items.  The application of one or more supervised learning methods to many embodiments of the present invention to mitigate the risk of such undesirable situations (as well as the consequences thereof) is both a necessary guarantee. and desirable.  An essential aspect for applying a supervised or even semi-supervised learning approach to the module (M03), rather than an entirely unattended approach in particular, lies in the ability of these modules to provide important safeguards. which, in turn, enable experienced users and / or administrators of an embodiment of the invention to maintain and control the performance of the embodiments of the present invention.  In some embodiments, one or more aspects relating to one or more modules or sub-modules described herein may be subject to the principles of processing within a database and / or implemented via one or more known varieties of these last principles.  Such database processing may be particularly desirable given the iterative and frequent transformation of content, especially when content is frequently added, deleted, and modified, or when aggregation of various content elements is expressly established by a human operator.  In one embodiment of the present invention, the weightings of product attributes generated by the text analysis module (M03) can be in any form and format intelligible to the matching module ( M09A), described in greater detail herein.  By way of non-limiting example, these can be expressed, on the one hand, by means of a simple value-attribute system, or on the other hand be represented by one or more matrices of transfer variously complex.  The relevant thematic category groupings inferred and extracted by this module (M03) can be stored in a centralized database, or distributed among several specially designated sites.  Said category groupings may furthermore be generated, extracted, and / or stored, according to a nonlimiting enumeration, according to translations, synonyms, equivalents or other abstractions thereof.  These category groupings may also typically be words in one or more languages, although in some embodiments they may also be an alphanumeric string that represents any human natural language or other representation using any other system by which information can be encoded and / or decoded.  It will be understood, given the various functionalities described herein and arising from the various logical problems that it must solve, including, without limitation, classification, novelty detection, knowledge discovery, and data processing, the Text analysis module (M03) can be implemented by way of an artificial neural network.  CLIENT ATTRIBUTES [0042] For several embodiments of the present invention, the presence of one or more sources of client intelligence is the necessary and critical counterpart of the text analysis module (M03) described in FIG. present.  The Customer Attribute Module (M13) can be thought of as a structure that merges this customer intelligence and makes it available to the system described herein in accordance with the stated purpose of the customer.  [0043] Client attributes are important aspects of the system in that they provide key customer intelligence data needed to develop effective marketing campaigns.  These campaigns ideally and optimally combine advertising content with the known interests of customers.  For the purposes of the embodiments of the present invention, collective trends can be collected and provide useful information for the management of the campaign in aggregate form; However, in its most valuable form, such customer intelligence provides a personalized insight into the experience and behavior of individual users.  Managers and other individuals designated as having sets of specific operator privileges compatible with the assigned task structures, the confidence level, and the objectives of the campaign to name a few non-limiting criteria, can be done. providing access to specific portions of the backbone infrastructure of an embodiment of the present invention, including, but not limited to, specific undisclosed URLs of administrator or operator web portal by via a secure connection to the main computer system on which an embodiment of the present invention is deployed.  Allowing such access to designated human operators accomplishes two vital objectives; first, it allows for the creation and management of advertising campaigns, and second, it ensures that all the modules of an embodiment of the present invention to which these first have access work properly.  Front-end access to embodiments of the present invention may follow a similar paradigm; however, it will be understood that some differences exist logistically and functionally exist.  A first consideration relates to the method by which embodiments of the present invention may obtain or even acquire users or customers to solicit through targeted advertising campaigns.  Such an acquisition and the subsequent interaction may take place by various means in various embodiments of the invention.  In some embodiments, the individual customers may create their own accounts in said embodiment through a web interface similar to the user account creation interfaces seen on various web sites, allowing said clients to subsequently connect to the system. ; non-exhaustive means to achieve this goal may include connecting to a dedicated user session by means of a secure connection of a web browser.  Indeed, these sessions could be established by the direct and explicit visit of a centralized website and / or server where an embodiment of the present invention would be hosted, all the interactions taking place at said site. Web following authentication of the user, which, in some embodiments can be completed, according to a non-limiting example, by entering a combination of a username and a password.  In other embodiments, the interaction with the system may be much less centralized and may employ technologies such as third-party persistent web cookies or other session-specific tracking technologies deployed on multiple external Web sites. both to target account-holding customers and to convey the behavior of customers to an embodiment of the present invention.  This latter scenario, while offering a less constrained user experience, provides an opportunity to receive targeted advertising solicitations when browsing Internet pages as part of its typical web browsing habits.  These external websites may be subject to affiliation either vague or narrow, but not necessarily to be administered by the same parties as the one that administers one or more embodiments of the present invention.  In some embodiments, advertisements displayed to persistently authenticated users may be broad in scope or may be limited by certain thematic criteria; For example, a camp site may choose to display advertising content exclusively related to camping equipment or to waive any thematic restrictions on advertising.  In other embodiments, the user-user and not the operator of a third-party website may be given the opportunity to decide whether or to what extent to limit any external content.  In some embodiments, once a client properly performs the logoff procedure of said embodiments, the tracking and related data collection may be terminated.  In another embodiment, advertising campaigns may be deployed for individual recipients via an email message in which the explicit and / or targeted advertising content is contained, integrated, or made accessible to a user, in any case or in part, through access to said linked content from said email message.  The tracking of clicks is the main method by which the preferences of each customer can be collected and tracked to then build his profile of interest.  In some embodiments, such as when embodiments of the present invention are used on computing platforms such as mobile devices on which a mouse click itself is not the means by which to select elements of an interface graphic, the appropriate modifications (such as touch screen pressure or the pronunciation of a voice command, for example) must be inferred.  When collecting information associating the occurrence of a click of a user for one or more advertisements viewed by said user, several embodiments of the present invention may also complete and correlate the profile of the user. interest of a user with less explicit and more passive session data.  By way of non-limiting example, information such as the amount of time spent watching a particular advertisement, the number of times the same ad was viewed, whether or not products belonging to families or categories of products Similar (and if yes, which) were seen, the platform on which the initial connections and subsequent sessions took place, the date, time and location information when available can additionally be collected and added to the data of attributes (M13) collected for the purpose of inference and the deduction of the client's interest.  In some embodiments, a machine learning algorithm of at least some features similar to those described in the text analysis module (M03) may be deployed to facilitate the conclusion of certain inferences specifically related to the data of customer attributes.  In still other embodiments, additional information collected may include payment patterns and language preferences, and may combine collected customer intelligence data with all the information requested and provided by the customer at the time of the inquiry. creating his account, such, his name and other demographic information such as his gender and age.  In most embodiments, the collection of customer information is performed for the purpose of generating weighted statistical data in a manner and according to a complementary process in some respects similar to that generated by the analytics module. text (M03).  In parallel, the statistical data representing, for a set of clients, a weighted and quantified ranking of the affinity of each of said clients with various relevant thematic category groupings extracted by the text analysis module (M03), is collected and calculated on a continual basis.  Said statistical data calculations are deduced from sources of behavior obtained, a non-limiting enumeration of which is provided herein by way of non-exhaustive examples.  In some embodiments, relative weighting, preference, or offset value may be assigned to one or more of said sources of behavior; these preference shifts may be the result of heuristics, an objective, or a simple editorial decision, or for any reason deemed appropriate by a system administrator.  In addition, the frame of reference against which the statistical data of a customer is calculated - and, in some embodiments, kept and stored - can, in various embodiments, be based on the customer's data profile itself. same, or be normalized to a set or to all clients whose attributes (M13) are determined.  Weighted statistical data produced by both the Text Analysis Module (M03) and the Client Attributes Module (M13) will then be matched (M09A) in a process to accurately solicit a customer with one or more advertising announcements. products using knowledge of which products and product categories are likely to be attractive to those customers.  Although the explicit interface and the presentation may differ according to the mode (s) of realization under study, the overall procedure remains essentially comparable, proceeding by iteration for one, for some, or for all the customers to both, said customers having affiliated with an embodiment of the present invention.  The decision as to whether the selection of target recipients according to the advertising content or vice versa can be implemented as complementary functionalities of various embodiments of the present invention, with one or both of the functions made available as an element of consideration specific to the specific situations faced by the administrative staff responsible for the planning and deployment of a particular marketing campaign.  Finally, in a manner analogous to that described for the text analysis module (M03), the client attribute information (M13) collected and described herein can be collected, stored, and represented by several means, including in some embodiments by maintaining, as a non-limiting example, an attribute-value table, and in others, through a more comprehensive representation of the client attribute data. using, for example, more complex transformation matrices.  Further discussion of the role of instrumental descriptive analytics in the embodiments of the present invention is further provided in the present disclosure.  PAIRING MODULE The matching module (M09A) receives the product attributes of the database (M01) and the weightings of the product attributes of the text analysis module (M03).  In addition, the matching module (M09A) is configured to match the attributes related to the previously described products and the individual client attributes (M13).  The extent to which product-related attributes are mapped to customer attributes (M13), including functional extraction considerations and the integration rate of the respective corpora, as well as any additional aspects described in present, are the subject of administrative considerations which can be envisaged for each embodiment, as authorized aspects or functions available from classes of different embodiments.  Nevertheless, the product-related attributes described above are adapted to produce a weighted set of client-specific product affinity data, a more detailed description of which is set forth herein.  An important aspect that governs the use of the matching module relates to the determination of which of the two types of main attributes (c. - -30-  linked to the customer or related to the product) on which to operate, and which of them will serve as a predicate.  More specifically, it must first be determined whether the purpose of a particular advertising campaign is to match a set of products with known or likely to be most receptive customers to those products, or vice versa, of matching a given set of customers with known or potentially most attractive products to those customers.  Although a significant proportion of the embodiments discussed herein presume this latter scenario, it will be understood that in many practical cases the first scenario may also prevail.  In such cases, a reading of this disclosure should be made, with the necessary modifications.  The nature of the inference and / or descriptive statistical methods involved in generating relevant thematic category groupings underlying the various embodiments of the present invention remain similar.  A key aspect of the step of the matching module is the integration of the attributes generated independently by the text analysis modules (M03) of client attributes (M13).  It will be appreciated that in various embodiments of the present invention, the various constituent elements, including the aforementioned modules, may be implemented and deployed on more than one computer and / or at more than one physical site.  At the same time, but distinct from these considerations, there are several other aspects related to variations in extensibility and interdependence of the technologies underlying the modules described herein that must be taken into account.  [0055] From one of these considerations, a large amount of data generally requires processing in a typical embodiment and in the deployment of the present invention.  Especially in light of the scale of processing dictated by massive data resources and assets - already considerable and expected to grow exponentially continuously - it is imperative, particularly in On the larger scale of the present invention, the time required to perform said processing does not constrain the operation, in particular, of the matching module (M09A) or its tributary modules, in particular the text analysis modules (M03). and customer attributes (M13).  Therefore, many of the technologies solicited to solve the various challenges introduced by large data can be considered and deployed, conveniently at both scale and embodiment, to fruitfully manipulate and efficiently handle large quantities. data relating to the various embodiments of the present invention.  [0056] In this respect, advanced parallel processing technologies emerge as necessary elements in both statistical corpora, one of which is customer-oriented and the other is product-oriented - to be matched.  To perform such matching, the matching module (M09A) determines, for each client in a given set, a statistically quantized affinity for a corresponding set of products whose content exists in the database.  In some embodiments, the architecture schemes, including but not limited to extraction-transform-load (ETL) may be implemented in the pairing module (M09A) to facilitate this process.  In addition to the above, data mediation principles and in particular ontology-based data integration forms may be variously incorporated into embodiments of the present invention in order to achieve this objective.  The result of the aforementioned matching process is the production of a client-specific product affinity data set associated with a set of clients, specified previously, to be targeted during a particular campaign.  Said data constitutes a quantified predictive measure corresponding to the extent to which it would have been calculated, for each customer among said set of customers - given the available behavioral histories - would have a particular interest for a given set of products selected from the advertising content base (M01) for purposes of a particular campaign.  In some embodiments, the maximum number of products to be retained for the purposes of a given campaign may be specified by a system administrator or a campaign administrator; in other embodiments, the assembly may incorporate the integral of the available content (M01).  In addition, the set of customer-specific product affinity data produced by the matching module (M09A) may be expressed in any scheme and format deemed appropriate for the given embodiments of the present invention; to cite two non-limiting examples, these may range from a simple comma-separated list of values to collections of large matrices to billions of entries.  Once these data have been calculated for all the desired customers, they are provided to the matching module and template layout of selected products (M09B).  CONCORDANCE MODULE [0059] The concordance module (M09B) is the main final component of the invention; its primary operations include the filing and selection (M09B1) of content or articles for solicitation purposes, and completing (M09B2) a specified advertising template with personalized content (including but not limited to advertising) ) for a given customer.  The success of the above two steps is the result of various aspects of the original content or the transformation and integration steps described above.  As an online and above all direct marketing solution, embodiments of the present invention are particularly influenced by the challenges and trends encountered in the field of variable data printing.  A key aspect in this enterprise is the cooperation between, on the one hand, the validation and ranking of the relevance of the personalized content to be presented (M09B1), and on the other hand by ensuring that said content fills and either displayed in a fairly rigid pre-specified layout or a template (M09B2).  The two submodules contained within the concordance module (M09B) cooperate to realize the operation of this paradigm.  It will be understood that even if the advertising content base (M01) can store many types of content, each potentially in different formats, generally, only a subset of such content appears in a template. unique personalized advertising for a particular client.  This is all the more relevant given the fact that various associated content instances may exist in the database (M01) for a given group of closely related products - or for nearly identical products.  For purposes of disclosing embodiments of the present invention, relevant information about a product to appear from the space allocated in the template and visible to the end recipient client is known as meta content of product attributes. .  In addition to receiving one or more sets of client-specific product affinity data described above, the matching module (M09B) also receives the meta-content of the product attributes of the advertising content base (M01) directly.  By way of non-limiting example, assume that an advertising campaign for a large-area electronics store should be coordinated by one embodiment of the present invention.  The matching module (M09A) could determine that the client attributes of a particular client (M13) suggest a considerably high level of interest in specific varieties of mobile phones.  Suppose further that the available client attributes were granular to the point of allowing predictive identification of a very specific model of mobile phone with which to solicit said client.  In such cases, customer-specific product affinity data for not only mobile phones, but for particular models of mobile phones could be transmitted to the matching module (M09B1) of the concordance module ( M09B).  In one embodiment of the invention, the mobile device model is predictively selected (e.g. -to-d.  the model having the highest quantified statistical compliance ranking for said client due to the existing client attribute data (M13) for said client) so that it appears on the template to be populated accordingly (M09B2), then generated and sent to the target client.  As part of the module's disposition functions (M09B2), the product whose appropriateness would have been determined for solicitation purposes from the customer is retained for display purposes in an advertising template layout, discussed in greater detail in present.
[0004] However, for a particular handset model, the content to be collected from the database (M01) and to be provided to the matching module (M09B) may include the device name, model information, a subset its characteristics (usually limited to the most prominent), the selling price of the device, and a product image. On the other hand, even if it were present in the back-end database (M01), contents such as the serial number used by a third-party manufacturer to designate an obscure electronic component in the handset, or perhaps the date the handset assembly could, for obvious pragmatic and contextual reasons, not be included in the custom and populated front-end template; similarly, the device's user manual, even if available (M01), would not normally be reproduced or displayed to the customer in such a scenario. In a variant example, the client attributes (M13) collected for a given client may suggest not one, but rather a plurality of mobile devices (having comparable or potentially different characteristics) whose descriptions are contained. in the database (M01) could be predicted to be of interest to the client. In this case, the mobile device models to be selected are selected (M09B1) according to their respective predictive rankings, and are provided to the stand submodule according to the ranking (M09B2). The latter sub-module (M09B2) accordingly fills the template with the subsets of mobile phone models, according to the respective classification assigned to each handset model as well as to any specified external selection criteria (eg to populate the template with a maximum of two mobile phone devices, placing the highest ranked device in the populated template first), to be discussed currently. Another equally important aspect of the concordance module (M09B) concerns the content management functions by which the module allows administrative users of an embodiment of the invention to choose from one or more generic advertising templates and their related graphical layout. While this content management feature typically includes aesthetic and design considerations, it also allows an administrator to make basic editorial decisions about content, including critical judgments about whether and how to prioritize, rank, and to coordinate the different aspects of individual advertising campaigns to be carried out. These functionalities include, as a non-limitative example, the possibility for one or more administrators to specify, to reach, or to optimize the specific objectives of artistic or economic nature, such as by modifying the relative positions of the selected articles of theirs. origin and their automatic positioning would have been conferred by the module (M09B) for one or more clients, as well as the possibility for one or more administrators to manually bypass the decisions made by the matching module (M09A) and the concordance module (M09B), removing and optionally replacing one or more of the products listed and arranged in a template populated automatically with one or more other elements of the database (M01). In various embodiments of the present invention, it is possible to have a library of not one but several templates or pre-developed layouts, each of which provides various possibilities for layout and customization of the appearance of the content. This customization may include, but is not limited to, font selection schemes, text size, image size, color scheme, corporate brand, and seasonal publication. Such decisions may be externally specified and provided to the matching module (M09B), or may be implemented as an additional interface component or sub-module within the concordance module (M09B). The output of the concordance module (M09B) is the specific template populated for the targeted client, which population results from the combined operations of all the modules described above, and which in many embodiments is customized and deployed according to the preferences. communicated to a particular customer in addition to other potential administrative or editorial considerations relating to the campaign itself. The template thus filled can be sent to the recipient via any distribution means, which may vary according to the embodiment. These populated model instances can provide either reformatable or non-reformatable content. In a non-limiting example, delivery means may include an e-mail message, wherein the body of the e-mail message addressed to a given client-recipient contains the populated template itself. In another embodiment, the custom populated template may be included in the same email message as an attachment in any format, including but not limited to a PDF file, or simply include a hyperlink to a designated web address. where a 5 tel content can be viewed and displayed using a web browser, such as in a site equipped with e-commerce features. In some embodiments, the custom template delivered and accessed by the recipient customer may be partially or fully loaded with tracking technology components described above, given the appropriate adaptations for each delivery mode. Once the content mentioned above, regardless of the distribution means, is made available to a client, said client may, in some embodiments, have the ability to access additional information on the subject. any or all of the elements appearing in the custom template by any means of selection, including, but not limited to, pointing, clicking and / or pressing the occupied model area by a particular advertisement of the product. In many embodiments, details of the nature of each attempt a client would have made to access more product information can be retransmitted to the customer's attribute (M13) module, using tracking technology elements described herein, in order to continually update, improve and improve the individual profile of said customer for future use, such as in a subsequent campaign. In an alternative embodiment, the information thus collected can also be used for analytical purposes beyond those directly related to that client; for example, data of interest for multiple clients may be aggregated and used for qualitative identification of features and trends of interest, as well as for varying valuing such interest among one or more clients as well as among or several articles presented to them. Indeed, in various embodiments and contexts, these assessments can be more concrete and quantifiable based on existing analytics, which can in turn be very useful in market studies; in other scenarios where analytics data are only partial or still entirely absent, evaluations may be useful for preliminary purposes, or hypothetically designed to provide an approximation of more than one. simple approximation. Finally, in some embodiments, the possibility of selecting one or more items from a populated template and provide a means by which to buy and / or obtain additional information, such as through a digital showcase, can also to be considered. PREDICTIVE ANALYSIS [0068] It will be understood that predictive analysis plays a central role in the function and operation of various embodiments of the present invention. It will be understood here that the purpose of predictive analytics can be appreciated in terms of three interconnected marketing optimization objectives. The first of these goals is to increase the click-through rate by ensuring that the relevant solicitations are targeted as accurately as possible, which means that they are issued to a known set of customers most interested in receiving them. This interest is the result of careful customer profiling efforts and is therefore indicated by a positive customer-recipient reaction, such as by clicking on a given advertising solicitation leading to additional summary information, such as the display. a trailer or preview of a video, the reading of the abstract of a publication or the headline, the caption and summary of a news article. The second objective is to increase the conversion rate, whereby said preliminary interest changes from a simple customer interest report to a completed transaction, which can be considered a commercial continuation of the first objective, said second objective being generally but not exclusively appreciated as the purchase of a requested good or service. The third and last objective is a kind of recursive optimization of its predecessors, but extends over the period of each relationship formed between a set of solicited clients and an embodiment of the present invention. The effectiveness of marketing is increased by these three objectives in two ways. The first of these includes establishing a holistic view of each individual from a given set of potential or targeted customers (M13), including the accumulation and storage of attributes of said customers. The second objective is to use the matching module (M09A) to arrive at exact predictions by which one or more items to be potentially solicited are paired with known client attributes (M13) so as to maximize the possibility that said solicitation is favorably or positively by the target client (s). [0070] Therefore, the practical delimitation of such "favorable reception" can be appreciated as a relative and fluid notion, which varies according to non-limiting factors such as the context in which an embodiment is deployed; moreover, the nature of the article itself may play a role in the interpretation of the reception that will be received by a targeted recipient. In the latter scenario, the notion of acceptance may vary from one click, screen pressure, or other means of selecting a solicitation within a template to the reading or viewing content in whole or in part, placing an online order for a particular property, for example, purchasing a particular service. Similarly, the related term "favorable appeal" may usefully describe these phenomena, especially when it comes to considering recipient responses from a positive and internal point of view or for statistical analysis. It will be further understood that the predictive aspects participating in the aforementioned coupling step (M09A) may vary. In particular, such predictive aspects may include, but are not limited to, the mere solicitation of a particular transaction, or the completed issue or successful delivery of a particular item to specific recipient customers or the delivery of a particular item to specific recipient customers. targeted based on a profile constructed of detailed preferences of said customers. The aspects involved in the predictive facet that can be understood by the embodiments of the present invention may also vary. In view of the rather broad potential scope of the present invention, these aspects - some of which are listed and described herein in a non-exhaustive manner - should nonetheless be appreciated as cumulative and even overlapping. It will be understood that the precise overall, extent and conceptual overlap of the customer attributes (M13) to be collected, stored and analyzed may vary depending on the overall marketing needs and / or the specific performance objectives of the mode of delivery. Embodiment of the envisaged invention. Similarly, the precise set of attributes to be collected and on which the predictive analysis evaluations are made by embodiments of the present invention are based may vary. In many embodiments, these analytics attributes, as well as the explanatory variables they encompass, may be specified by a human operator having administrator and / or campaign management privileges. In addition, the overall extent and range of these attributes may be specified, without limitation, on a basis that is either customer-specific or campaign-specific. The set of attributes actively involved in the predictive analysis operations carried out by embodiments of the present invention may further be selected from a set of potentially larger explanatory variables, which variables the embodiment of the present invention would have. has been configured to follow for each client profile it builds. Embodiments of the present invention may further benefit from a modular design by means of which the predictive analytics components - be they specific attributes to collect customer profiles or even means by which algorithmically treat these attributes - are modifiable following the deployment and the active use of said embodiments, these changes having in turn a minor negative effect or minimal on the proper operation of this embodiment. In a manner similar to the modifications and / or upgrades of any module described herein, such modifications may be variously deployed, if required, by various strategies, and may include on-site deployments, in situ, remotely or by the eligible administration personnel, any deployment being compatible with the optimization objectives described herein and with the particular requirements arising from the needs of the embodiment of the invention. These modifications may vary in scope and may be stand-alone, as in a self-executing software upgrade program, or may require extensive structural changes to an embodiment of the present invention. Similarly, these changes can be deployed regularly, periodically or when campaign management deems it appropriate or necessary, and can be motivated by innovative analysis strategies, or to optimize those already in place. The moment or the propitiousness of an implementation of a modification can come when it is appreciated that these components, either specifically or broadly, require said modification to advance the objectives of carrying out the modification. present invention. Typically, we come to make such decisions jointly following a careful examination of the sub-optimized facets of the mode (s) of realization and a collaborative consultation on the tangible remedies to be applied to said sub-optimizations by the administrative staff of the marketing campaign and one or more persons skilled in the art and who have the ability to develop such changes. POSSIBLE ENTRY VALUES [0075] A non-limiting discussion of the type of information collected and which partially fulfills the set of client attribute data (M13) currently follows. Each of the following input values, which is enumerated herein without limitation, may contribute to the predictive analysis processes used by various embodiments of the present invention to optimize the targeting and matching of recipients to the solicited items. . Although the location of a given target client can be disclosed statically - for example in a profile creation form - when said client is affiliated with an embodiment of the present invention, it may be considered desirable to exploit more precise information on the geographic location of said client for the purpose of predictive analysis operations. For example, the GPS coordinates, the specified timestamp values, and basic device and platform information upon which a commercial solicitation is received may be requested by an embodiment of the invention, which may track and combine the data as well as the viewing habits of the target customer. This information can be used to draw conclusions about the probability of success with solicitations of the same or similar nature in the future based on the location of a potential customer. Conversely, predictive analytics can also be exploited to discern possible patterns in the types of articles whose solicitation is successful based on location, device, and timestamp data. A second type of possible input value which may be statistically useful is the number of articles presented to a given customer as part of a particular marketing campaign or even a solicitation message or a message. template given. This information may be combined with other data described herein to evaluate and possibly optimize successive marketing initiatives. These optimizations can focus on an adaptation of the nature, duration, and presentation of future correspondence to be transmitted to a target customer. A third type of possible input value could include parameters intended to quantify the relative magnitude of the traffic covered by an embodiment of the invention. These can include, but are not limited to, the number of active clients to whom solicitations can be sent, the number of solicitations sent to each customer (or a group of customers) as part of a specific campaign, or the number of total items from one or more sources (eg, retailers, distribution companies, publishers) available or posted to all such customers. A fourth type of possible input value could focus on aspects specific to conversion rates or click rates achieved through a given embodiment of the present invention. For example, statistics such as counting clicks (or some other form of favorable acceptance as described herein) made by targeted customers as a result of a given solicitation may be useful for this purpose. . An associated value could even count down the number of favorable acceptances of all the links displayed on all the websites during a specified time interval. Similarly, another value could total the number of successful transactions (eg a solicited or a service purchased by a target customer) connected by an embodiment of the invention during a given time interval. A fifth type of possible input value follows the theme of the preceding type, but from a point of view featuring each individual customer profile, as opposed to a cumulative total collected integrally from all sources taken together. . Thus, in the present grouping, parameters such as the total number of favorable acceptances made by each specific client profile during a given campaign could be counted. In a related manner, the total number of favorable acceptances collected by a particular customer profile holder for a specific merchant, for example, could be collected. Finally, an aggregate of the total number of transactions made by each account holder could also be collected. In addition, the attributes and data collected and stored in one or more explanatory variables may, in some embodiments, be otherwise aggregated or combined. For example, the data stored in several different explanatory variables within the client attribute profile of a specific client can be combined to derive one or more predictive inferences about the client. Conversely, the data collected and stored for the same explanatory variable and belonging to a set of clients can also be combined to derive one or more predictive inferences on a given group. Finally, it will be understood that for an embodiment of the invention to come to construct a global vision of a set of target customers and to draw more sophisticated predictive conclusions following the realization of better optimized and targeted campaigns. more complex combinations of explanatory variables can be used by concurrently aggregating the aforementioned strategies for combining client attributes, whether for multiple variable cases / single client or single variable / multiple clients. It is possible to exploit types of input values such as those described above as well as the predictive analyzes that result therefrom to come, in various commercial scenarios, to results while respecting the main objective of achieving the present invention to increase the effectiveness of marketing. For example, a chain of pizzerias having a local franchise could use an embodiment of the present invention or retain the services of a company using the latter to solicit a given customer, which said embodiment knows he prefers to eat pizza on certain key days of the week and especially at certain key moments of the day. [0083] Similarly, a casino that offers players a personal authentication mechanism (such as a magnetic strip or a smart identity card with centralized integrated payment functionalities) to facilitate the game could thus benefit from such a system. analysis. The casino could, for example, identify patterns of behavior of a given player and ensure that casino employees take specific actions, either to trigger particular behavior (such as playing a specific game or buying a drink or meal) or to prevent certain scenarios or to prevent certain outcomes from happening (for example, by intervening with a player whose gambling profile suggests that he or she is a low-income person who may have a problem with gambling. that she stops playing or that she even considers a self-prohibition). In another scenario, the relative placement of articles or descriptions in a commercial solicitation can be combined with the additional predictive analysis according to a specific marketing objective. For example, placement of item descriptions in a populated and customized solicitation template for a specific customer may be subject to a condition that the placement adheres to specific criteria. For example, said articles may be arranged so that they appear so as to encourage a target recipient to first browse less interesting (but nevertheless desirable) articles. As an adjunct, the placement of items may follow another criteria, such as placing items at low or high product profit margins in key locations in the template, or items with higher popularity ratings in places so-called "privileged" within a template. In addition, each of the preceding data values enumerated in a nonlimiting manner may be combined in data mining and analytical exercises for the purpose of extracting solid results with ease and in a manner that is easily adaptable. Thus, it will be understood that the precise combinations of the explanatory variables, as well as their usable inferential value, vary according to the needs of the specific embodiments of the present invention. POST-CLICK ANALYSIS [0087] An important element in the collection of the predictive analytics described herein involves the iterative process by which the post-click tracking data is communicated to the various modules of the present invention upon receipt by a client of a populated template. As described herein, acceptance refers to a general signaling of some interest by a customer for one or more items in the template that it receives. Various scenarios can be imagined in which a customer generates a favorable acceptance for one or more elements of a populated template; similarly, in other cases, a client may signal some superficial interest by viewing all the elements of said populated template.
[0005] In one embodiment of the invention, a portion of the predictive analytics collection process may be configured to interpret not only the post-click data collected on the relative interest of a clientele in a party, or all the requested content, but also in order to be able to interpret the relative interest level for the products by measuring, for example, the time spent in relation to one article compared to another contained in the populated template. In at least one embodiment of the invention, such an attraction could thus be measured by means of one or more input values specified and configured to quantify the client's interest as a function of the time that it passes to investigate each solicited item in a populated template. The signaling of acceptance by a client of an article published in a populated template operates a feedback mechanism involving two main modules described herein, namely the text analysis module (M03) and that of the attributes. customers (M13). This is because the acceptance is considered, on the one hand, as a partial guarantee by the customer of the set of keywords generated by the text analysis module (M03) for a particular article. . On the other hand, the customer's interest in said article can also be included in one or more predictive analysis input values without any link with keywords or even with the attributes of the article alone. On the other hand, the interest for a particular article can be considered as the natural extension of a behavioral trait which can become more and more obvious following successive campaigns as well as at the end of associated collections of elements of post-click predictive analysis. [0090] Based on an example developed above, it may be that it is known that a customer has a preference for a particular model of mobile phone. However, the predictive analytics collection for said client may also allow an embodiment of the invention to have an input value type configured to determine the color preferences that the client would have for articles. solicited. Therefore, in such a scenario, the realization of the invention does not know that said customer prefers articles of blue color, as opposed to those of red color, which is however the case. In addition, in this scenario, the populated template viewed by said client includes entries for two variants of said mobile phone model which differ only in their respective color, one of the requested variants is blue and the other red. In this scenario, a favorable acceptance of the blue device model could result, while the red model would be completely dodged by the customer. In this case, a quantitative increase of the favorable acceptance in said customer's scores for the subset of the attribute weightings corresponding to the mobile phone will be increased, in addition to said customer scores corresponding to the input value for the blue color. . In one embodiment (Figure 2), at such a favorable customer-generated acceptance, the text analysis module (M03) may be requested to identify article attribute weightings for the telephone. mobile having generated said favorable acceptance, and to provide them to the customer attribute module (M13). In addition, the customer attribute module receives click track data as an input value configured to identify color preferences and progressively increase the input value identifying a preference for color. blue. [0091] Finally, it will be noted that the implementation of the click tracking process and the collection of predictive analysis data described herein should be assessed in the broad and non-limiting sense and therefore not be restricted. or limited to only the embodiments and examples developed for the sole purpose of illustration.

权利要求:
Claims (10)
[0001]
REVENDICATIONS1. An article classification and selection system comprising an advertising content base containing a set of texts for a series of articles; a text analysis module configured to accept said article descriptions and to generate weighted article attribute values extracted from said article descriptions; a pairing module configured to pair said item attributes and said weighted item attribute values with the client attributes to produce a weighted set of customer-specific article affinity data.
[0002]
The system defined in claim 1, further comprising a match module and template arrangement of selected products configured to accept said client-specific article affinity weighted data of said pairing module, at least a portion of which said article descriptions extracted from said content base, and contains one or more of said article attributes, or one or more of said article attribute weightings, and externally provided template and matching criteria for populating said Template with custom content and layout for a customer with the client attributes.
[0003]
The system of claims 1 or 2, wherein said matching module and template layout of selected products is configured to populate a content template in which a subset of articles from said content base is presented. 51 -
[0004]
The system defined in claims 1, 2 or 3, further configured to externally substitute one or more template stand choices made by the match module and template layout of selected items.
[0005]
The system defined in one of claims 1 to 4, wherein the text analysis module is configured to interpret, infer, and generate relevant thematic category groupings of said item descriptions and provide said groupings to the module. pairing.
[0006]
The system defined in claim 5, wherein the result of said inference and generation is one or more category groupings, where part or all of the literal content thereof can be expressed by synonymous expressions, translations, equivalent or external formulations of various abstractions to the information presented in said description of the article itself.
[0007]
The system defined in claim 6, wherein said thematic category groupings are expressed as alphanumeric strings.
[0008]
8. The system according to any one of claims 1 to 7, wherein the text analysis module is implemented by means of a neural network.
[0009]
9. System according to any one of claims 1 to 8, wherein the content template is an advertisement template.
[0010]
The system defined in one of claims 1 to 9, wherein said content and said personalized provision for said customer are used for mass solicitation purposes involving email or other internet marketing mechanisms.

类似技术:

---

公开号 | 公开日 | 专利标题

---

FR3016459A1|2015-07-17|

---

US10380650B2|2019-08-13|Systems and methods for automating content design transformations based on user preference and activity data

---

Behera et al.2020|Personalized digital marketing recommender engine

---

US8924265B2|2014-12-30|System and process for improving product recommendations for use in providing personalized advertisements to retail customers

---

Bari et al.2016|Predictive analytics for dummies

---

CN109543111B|2021-09-21|Recommendation information screening method and device, storage medium and server

---

US8108329B2|2012-01-31|System and process for boosting recommendations for use in providing personalized advertisements to retail customers

---

US8725559B1|2014-05-13|Attribute based advertisement categorization

---

US20070233622A1|2007-10-04|Method and system for computerized searching and matching using emotional preference

---

WO2009137152A1|2009-11-12|System and process for receiving boosting recommendations for use in providing personalized advertisements to retail customers

---

CN108140041A|2018-06-08|It is clustered for the viewing time of video search

---

Chen et al.2011|Social opinion mining for supporting buyers’ complex decision making: exploratory user study and algorithm comparison

---

US20090198553A1|2009-08-06|System and process for generating a user model for use in providing personalized advertisements to retail customers

---

Stevenson2016|Data, Trust, and Transparency in Personalized Advertising.

---

US20160188587A1|2016-06-30|Methods and systems for providing expert media content sessions

---

US20090198552A1|2009-08-06|System and process for identifying users for which cooperative electronic advertising is relevant

---

Jain et al.2017|Improving customer experience using sentiment analysis in e-commerce

---

US20210224858A1|2021-07-22|Systems, Devices, and Methods for Autonomous Communication Generation, Distribution, and Management of Online Communications

---

US20090198556A1|2009-08-06|System and process for selecting personalized non-competitive electronic advertising

---

Mitsis et al.2018|Social media analytics in support of documentary production

---

US20090199233A1|2009-08-06|System and process for generating a selection model for use in personalized non-competitive advertising

---

Elahi et al.2021|Recommender Systems: Challenges and Opportunities in the Age of Big Data and Artificial Intelligence

---

Adomavicius et al.2006|Personalization technologies: A process-oriented perspective.

---

Ma et al.2021|Image-mining: exploring the impact of video content on the success of crowdfunding

---

Purwaningwulan et al.2018|The uniqueness of influencer marketing in the Indonesian muslim fashion industry on digital marketing communication era

同族专利:

---

公开号 | 公开日

---

WO2015106353A1|2015-07-23|

---

US20160335674A1|2016-11-17|

---

CA2973706A1|2015-07-23|

引用文献:

---

公开号 | 申请日 | 公开日 | 申请人 | 专利标题

---

US20130282713A1|2003-09-30|2013-10-24|Stephen R. Lawrence|Personalization of Web Search Results Using Term, Category, and Link-Based User Profiles|

---

US20120066253A1|2010-09-15|2012-03-15|Cbs Interactive, Inc.|Mapping Product Identification Information To A Product|

---

US6460036B1|1994-11-29|2002-10-01|Pinpoint Incorporated|System and method for providing customized electronic newspapers and target advertisements|

---

US7383258B2|2002-10-03|2008-06-03|Google, Inc.|Method and apparatus for characterizing documents based on clusters of related words|

---

US8321278B2|2003-09-30|2012-11-27|Google Inc.|Targeted advertisements based on user profiles and page profile|

---

US20070027901A1|2005-08-01|2007-02-01|John Chan|Method and System for Developing and Managing A Computer-Based Marketing Campaign|

---

US20110264512A1|2010-04-26|2011-10-27|Microsoft Corporation|Using structured offer data to create keyword based advertisements|

---

US20120296991A1|2011-02-23|2012-11-22|Nova Spivack|Adaptive system architecture for identifying popular topics from messages|

---

US20130144605A1|2011-12-06|2013-06-06|Mehrman Law Office, PC|Text Mining Analysis and Output System|

---

US8818788B1|2012-02-01|2014-08-26|Bazaarvoice, Inc.|System, method and computer program product for identifying words within collection of text applicable to specific sentiment|

---

US11100523B2|2012-02-08|2021-08-24|Gatsby Technologies, LLC|Determining relationship values|

---

WO2014085908A1|2012-12-05|2014-06-12|Jonathan Michael Lee|System and method for finding and prioritizing content based on user specific interest profiles|

---

US8762302B1|2013-02-22|2014-06-24|Bottlenose, Inc.|System and method for revealing correlations between data streams|US20160314184A1|2015-04-27|2016-10-27|Google Inc.|Classifying documents by cluster|

---

US9904661B2|2015-06-23|2018-02-27|International Business Machines Corporation|Real-time agreement analysis|

---

US10445343B2|2015-07-16|2019-10-15|Interactive Data Pricing And Reference Data Llc|Systems and methods for data exchange and conversion|

---

US11069001B1|2016-01-15|2021-07-20|Intuit Inc.|Method and system for providing personalized user experiences in compliance with service provider business rules|

---

US11030631B1|2016-01-29|2021-06-08|Intuit Inc.|Method and system for generating user experience analytics models by unbiasing data samples to improve personalization of user experiences in a tax return preparation system|

---

US10621597B2|2016-04-15|2020-04-14|Intuit Inc.|Method and system for updating analytics models that are used to dynamically and adaptively provide personalized user experiences in a software system|

---

US10621677B2|2016-04-25|2020-04-14|Intuit Inc.|Method and system for applying dynamic and adaptive testing techniques to a software system to improve selection of predictive models for personalizing user experiences in the software system|

---

US10411946B2|2016-06-14|2019-09-10|TUPL, Inc.|Fixed line resource management|

---

US10783575B1|2016-07-01|2020-09-22|Apttus Corporation|System, method, and computer program for deploying a prepackaged analytic intelligence module for a quote-to-cash application while protecting the privacy of customer data|

---

US11205103B2|2016-12-09|2021-12-21|The Research Foundation for the State University|Semisupervised autoencoder for sentiment analysis|

---

TWI644272B|2016-12-29|2018-12-11|臺灣中小企業銀行股份有限公司|Shopping guiding system for digital bank based on big data and method thereof|

---

US20180253677A1|2017-03-01|2018-09-06|Gregory James Foster|Method for Performing Dynamic Data Analytics|

---

US10943309B1|2017-03-10|2021-03-09|Intuit Inc.|System and method for providing a predicted tax refund range based on probabilistic calculation|

---

US11232508B2|2017-04-11|2022-01-25|Apttus Corporation|Quote-to-cash intelligent software agent|

---

US10937078B1|2018-01-31|2021-03-02|Amazon Technologies, Inc.|Dynamic generation of item description content|

---

CN108694160B|2018-05-15|2021-01-22|北京三快在线科技有限公司|Article generation method, article generation device and storage medium|

法律状态:
2016-01-26| PLFP| Fee payment|Year of fee payment: 2 |

---

2017-01-26| PLFP| Fee payment|Year of fee payment: 3 |

---

2018-01-26| PLFP| Fee payment|Year of fee payment: 4 |

---

2020-01-28| PLFP| Fee payment|Year of fee payment: 6 |

---

2020-03-20| PLSC| Search report ready|Effective date: 20200320 |

---

2021-04-30| RX| Complete rejection|Effective date: 20210325 |

优先权:

---

申请号 | 申请日 | 专利标题

---

US201461927723P| true| 2014-01-15|2014-01-15|

---