WO2018015515A1

WO2018015515A1 - Methods for opinion sharing, computer programs and hardware for implementing methods

Info

Publication number: WO2018015515A1
Application number: PCT/EP2017/068426
Authority: WO
Inventors: Thomas Moreau; Charles THIERY
Original assignee: Trendstag
Priority date: 2016-07-21
Filing date: 2017-07-20
Publication date: 2018-01-25

Abstract

Opinion sharing method implemented in a communication network comprising a server capable of communicating with one or more client devices, and comprising: storing of subjects in at least one database of the server, each subject being associated with at least one computer label and opinion data; following an identification, on a service platform used by client devices for communicating, of a computer label of a subject in the database of the server, implementing of means for collecting a vote or a comment and receiving by the server of a vote or a comment relating to the subject identified by means of the label of same; updating the opinion data associated with said subject depending on the received comment or vote.

Description

METHODS OF SHARING OPINION, EQUIPMENT AND COMPUTER PROGRAMS FOR IMPLEMENTING METHODS

The present invention relates to methods for sharing opinions, and to systems, equipment and software (computer programs) for implementing the method.

Hashtags and mentions are widely used on social media, especially in social media users' posts. Hashtags and mentions consist of a word or a group of words without space introduced by an invariant prefix. The prefix "#" is used to form a hashtag, while the prefix "@" is used to form a statement. Hashtags and mentions are types of labels (in English, "tag"), which are recognized and interpreted by software robots as metadata fulfilling their own functions.

In a publication, hashtags allow users to mark a reaction, an emotion or a subject, as the following examples show: "#Ready"; "#Happy"; "#BarackObama". Hashtags can be combined, used alone, or embedded in a sentence. For example, the following post has two hashtags: "#GoodPictureCamera: The #CameraMaker super camera that captures every detail of a painting."

A single hashtag can represent several subjects. For example, the hashtag "#Paris" can represent both the French capital and the city of the same name in the American state of Texas. Conversely, the same subject can correspond to several distinct hashtags. For example, the two hashtags "#USA" and "#UnitedStates" will correspond to the subject of the United States of America. Hashtags do not depend on any system.

On a platform, a reference refers to a user's profile, such as "@RichardDoe", which can be used on the Twitter® platform. A mention and a hashtag can refer to the same subject, such as "@BarackObama" and "#BarackObama". In a publication, mentions and hashtags can be combined, as shown in the following example: "Next Thursday, appointment to our 1st #meetup dedicated to fundraising with @TiuSystems".

The proliferation of hashtags and mentions began with Twitter®, then spread to Facebook®, Google + ®, and Instagram®, to drive their integration with search engine results. This has been technically made possible by the fact that hashtags and mentions use a prefix and do not contain any space, making them easily identifiable to a human being and a machine.

The ability of machines to isolate these types of tags has allowed hosting platforms to associate URLs with them. On social networks, the URL associated with a hashtag refers to all the publications of this platform containing the same hashtag, while the URL of a mention refers to the account of the user of the platform.

Natural language processing is a discipline on the borderline of linguistics, computer science and artificial intelligence. One of the many applications of this discipline is the computerized analysis of feelings based on data from text-based sources (opinion mining). The analysis of large amounts of data (in English, "big data") makes it possible to statistically deduce the general feeling of a community with regard to a subject.

Interest in opinion mining has increased with the development of social media (forums, social networks, etc.) that produce data in real time and in large quantities. Opinion mining has applications in the field of prediction and supervision, particularly with companies that have shown that they implement tools to retrieve information on their reputations, for example by using public data flows. presentations on forums and social networks, so as to oversee the feelings of their consumers to improve their communication.

The current issues of opinon mining include the need to develop algorithms to determine in a textual content, for example a sentence: the polarity of the feeling conveyed that can be positive, negative or neutral; the subject (s) targeted by the feeling; the meaning of the subject (s). This problem generates the sub-issues below.

Define a subject involves named entity recognition techniques that aims to isolate textual objects (ie a word, or a group of words) categorizable in classes such as: names of people, names of organizations or companies, place names, quantities, distances, values, dates, etc. Named entity recognition comes up against one of the unresolved problems in the natural language processing of lexical disambiguation which is to determine the meaning of a word in a sentence when it can have several possible senses, polysemous word.

Clearing the polarity of the feeling of a sentence as determining and defining the meaning of the subject (s) targeted by it can only lead to a probabilistic result limited by problems related to the processing of natural language which are among others that the meaning of a sentence depends on the context in which it is used, the type of language used, as well as the person who wrote it.

In the public data feeds exposed on forums and social networks, the mentions and hashtags contribute to remove the ambiguity of the words or the polysemic subjects that they mark because they are easily isolable and associated with data via their URL. For example, in the case of the identification of a polysemous subject, the fact that the URL of a mention refers to the page of the subject that it represents can make it possible to identify the subject that it represents.

In the case of (s) word (s) or polysemic subject (s) represented by a hashtag, the publications referenced on the same hashtag will not help to define with certainty the meaning of (s) word (s) or subject that the hashtag marks in its context but will allow through the analysis of the publications referred to calculate a probability that it is this topic or the meaning of this (s) word (s) ) rather than another.

In the collection of opinions, hashtags and mentions may help to remove ambiguities about identifying the subject of a content, but do not offer a secure solution.

The purpose of this disclosure is to improve the situation.

One object of the present disclosure is to provide opinion-sharing tools, opinion-sharing methods, and opinion-sharing equipment and software, intended in particular but in a non-limiting manner for users of the opinion. Internet network.

In one aspect, the present disclosure proposes a new type of computer tag, called "TrendsTags" or "OpinionTags", that allow their users to share an opinion on a topic related to the TrendsTag. The TrendsTags can advantageously be used on all types of social media, and are not, unlike hashtags and mentions, specific to social networks and some social media (such as forums on the Internet). For example, the TrendsTags can be used on social media such as any HTML page posted on the Internet or an Intranet, for example an article published on the Internet or a blog post .

The TrendsTags provide a tool designed to gather the opinions of users, by allowing different users of the network (eg the Internet) to share a polarized opinion on a given topic. In another aspect, the present disclosure provides several methods of using the TrendsTags, including methods for creating TrendsTag, sharing opinions about a topic associated with a TrendsTag, expressing a vote on an associated subject. a TrendsTag, and generating a comment on a topic associated with a TrendsTag.

With regard to a subject and at a given moment, the opinion of a user can indeed be positive or negative. An opinion will be polarized in the present disclosure if it is positive or negative. The voting and commentary features associated with the TrendsTags described below may provide polarized views of the subject that the TrendsTag represents. The TrendsTags and related services described herein provide an alternative to determining the opinion (s) emerging from a content by known computer techniques (which produce as discussed above uncertain results) by allowing users to provide Polarized opinion on an identified subject.

Thus, the TrendsTags allow a collection of opinions based on polarized opinions generated by users rather than a probabilistic calculation of them.

According to another aspect, computer programs are proposed, each loadable in a memory associated with a processor, and respectively comprising portions of code for implementing the steps of the various methods proposed by the present disclosure at the time of the invention. execution of said program by the processor, as well as data sets respectively representing, for example by compression or encoding, said computer programs.

Another aspect relates to a non-transient storage medium of a computer executable program, comprising a data set representing one or more programs, said one or more programs including instructions for executing said one or more programs. by a computer comprising a processing unit operatively coupled to memory means and an input / output interface module, driving the computer to execute the proposed methods.

The tools, processes, equipment and computer programs provided are particularly well, although not exclusively, for use by Internet users on the Internet or an Intranet. But they are also suitable for any communication network comprising at least one server node capable of communicating with one or more client nodes.

A first aspect of the invention thus relates to an opinion sharing method implemented in a communication network comprising a server capable of communicating with one or more client devices, and comprising:

storing subjects in at least one database of the server, each subject being associated with at least one computer tag and opinion data; following identification, on a service platform used by client devices for communicating, of a first subject associated with a first computer tag in the database of the server, reception by the server of a vote or a relative commentary the first identified subject;

update the opinion data associated with the first topic based on the vote or comment received.

According to one embodiment, the first subject may further be associated with a first web page and a first resource address in the database and upon receipt of a request by a client device including the first resource address, the method includes sending the first web page to the client device.

In addition, the first resource address can be inserted into a web page of the service platform.

Alternatively or in addition, the first web page comprises at least one voting function, and in which the vote is received by the server via the voting function of the first web page and in which the update of the data of Opinion associated with the first topic includes the recalculation of a score of at least one voting state for the first subject.

In addition, a positive voting URL and a negative voting URL may be displayed on the first web page, the positive voting URL including an identifier of the first subject and a positive voting status and the negative voting URL including the identifier of the first subject and a negative vote status.

In addition, the method may furthermore comprise, following the reception of a request comprising a positive voting state or a negative voting state:

generate a comment URL including a comment identifier and a vote identifier;

- address to the client device the comment URL allowing a user of the client device to make a comment;

- receive a request from the terminal including the comment formulated and the comment URL for storage in association with the first topic in the database.

Alternatively, the first computer tag may include a subject tag followed or preceded by a vote tag, the subject tag may include a prefix and a character string relating to the first subject and wherein the vote tag may include a prefix and a character string relating to a voting status, and the vote or comment corresponding to the first subject may be the voting state of the vote tag.

In addition, the first subject is identified from the character sequence by recognizing named entities.

According to one embodiment, the database of the server may be powered by third-party databases and the storage step may include receiving at least one subject, formatting the new subject so as to compare it with other topics of the server database, and curation of the new topic so as to avoid any introduction of duplicate topics in the server database.

According to one embodiment, the storage in the server database can comprise receiving a request for creating a subject, from a client device, and generating by the server a new computer tag in association with the new subject, for storing the new subject in association with the new computer tag in the database.

According to one embodiment, the method may include scanning web pages to identify the first tag, and upon identifying the first tag in a content of a given web page, the update of the opinion data comprises adding the content as a comment associated with the first topic.

A second aspect of the invention relates to a computer program medium storing instructions which, when implemented by a processor, cause this processor to implement the steps of the method according to the first aspect of the invention. .

A third aspect of the invention relates to a server capable of communicating with one or more client devices and with a service platform via a communication network, the server comprising:

at least one database storing subjects, each subject being associated with at least one computer tag and opinion data in the database; a communication module for receiving a vote or a comment relating to a first subject identified on the service platform used by client devices for communicating, the first subject being associated with a first computer tag in the database;

a management module configured to update opinion data associated with the first subject based on the vote or comment received.

A fourth aspect of the invention relates to a communication system for the sharing of opinion comprising a server according to the third embodiment of the invention and a service platform capable of communicating with said server and via which client devices are able to communicate.

Features and advantages of the present disclosure will appear in the following description of nonlimiting exemplary embodiments, with reference to the accompanying drawings, in which:

FIG. 1a is a diagram illustrating a client-server system on which the proposed methods can be implemented;

FIG. 1b is a diagram illustrating a client device on which the proposed methods can be implemented;

FIG. 1 c is a diagram illustrating a server device on which the proposed methods can be implemented;

FIG. 2 illustrates an example of a TrendsTag web page according to one or more embodiments;

FIG. 3a illustrates a Web page containing a TrendsTag, according to one or more embodiments;

FIG. 3b illustrates a window of information on a TrendsTag superimposed on a Web page containing the TrendsTag, according to one or more embodiments;

FIG. 4 illustrates an example form for creating a TrendsTag according to one or more embodiments;

Fig. 5a illustrates a data structure of a subject according to one or more embodiments;

FIG. 5b illustrates a data structure of a TrendsTag according to one or more embodiments;

In the following detailed description of embodiments of the invention, many specific details are presented to provide an understanding more complete. Nevertheless, those skilled in the art may realize that embodiments can be practiced without these specific details. In other cases, well-known features are not described in detail to avoid unnecessarily complicating the description.

The invention will be described below in the context of non-limiting users of the Internet network and social networks (forums, opinion sites, etc.) implemented on the Internet.

In what follows, the term "subject" will mean a semantic subject, representable by a word, a character or a group of words or characters in a language, and by a set of data, such as a string of characters, recorded in a computer memory. A subject could be a person, an institution, a community, a place, a product, a project, a company, a work, a show, an event, a social issue, etc., represented by a corresponding data set. . The following list provides some sample topics: "Windows 10"; "Edward Snowden"; "Abortion"; "Google" "New York," "Gay Marriage," etc.

The methods, devices, and systems provided by the present disclosure are described below in the non-limiting context of a client-server architecture, of which FIG. Figure 1a illustrates an example.

Figure 1a shows an exemplary system architecture (10) in one or more embodiments, including a plurality of client devices (11a, 11b) each operatively coupled to a server device (13). through a network (12), for example the Internet. The client devices (1 1 a, 1 1 b) and server (13) are configured to communicate by means of a set of communication protocols.

Depending on the embodiment, the client device (1 1 a, 1 1 b) can be implemented within a computer, a smartphone, a tablet, a laptop, or any electronic apparatus configured to implement some or all of the functions described below for the client device, and to communicate with a server according to a set of communication protocols.

Figure 1b shows a diagram illustrating some of the elements of a client device (20) adapted for implementing different aspects of the proposed methods.

The client device (20) comprises a processing unit (21), a memory (22), an internal communication bus (not shown in the figure) allowing the communication between the different components of the client device (20), a module of management of the input / output interfaces (23), and a screen (24). The processing unit (21) comprises one or more processors. The memory (22) comprises one or more memories, such as for example an internal memory as well as another medium for storing data (such as a SIM card, a removable memory card, or a memory connectable to the client device (20) ), whether internal or not to the client device (20). Depending on the embodiment, the memory (22) may comprise random access memory (Random Access Memory) and read only memory (ROM). The ROM retains in particular the boot software (25) of the client device (20). The RAM memory retains the operating system software (26) of the client device (20) as well as application programs or other software modules as well as data (27). It may in particular contain a TrendsTags management application program (28), configured to inter alia process the data exchanged with the server, and which may for example be implemented according to the embodiment chosen in the form of a dedicated application. or a program cooperating with an Internet browsing program, coupled to the processing unit (21). The memory (22) is thus able to store software program instructions executable by the processing unit (21), and in particular software program instructions which implement the functionalities corresponding to the different aspects of the methods described herein with respect to regards the client device. In the case of a software implementation, the program (28) implementing the proposed method can thus be stored in the memory (22). The processing unit (21) can execute this program (28) and thus implement these functionalities, while also controlling the various modules (display driver on the screen, input / output interface driver, etc.) which can be used during execution. The processing unit (21) is also coupled to an input / output interface management module (23), which is itself coupled to the memory (22) and to the various input / output interfaces whose device client (20) is provided, including the screen (24). The input / output interfaces with which the client device (20) is equipped may comprise, in addition to the screen (24), for example a keyboard, a loudspeaker, a microphone, a lens for shooting, an interface card network, or any other type of input / output equipment.

Figure 1c shows a diagram illustrating some of the elements of a server device (30) adapted for implementing different aspects of the proposed methods. The server device (30) comprises a processing unit (31), a memory (32), an input / output interface management module (33), a communication module (34), a user management module (35) ), a score management module (36), and a subject management module (37).

In the architecture illustrated in FIG. 1c, the memory (32), the input / output interface management module (33), the communication module (34), the user management module (35), the score management module (36), and the subjects management module (37) are each operably coupled to each other through the processing unit (31).

In one embodiment, the input / output interface management module (33) is configured to receive data (and, for example, requests) from client devices, and to transmit the received data to the processing unit. (31) for performing treatments according to one or more embodiments described herein. The input / output interface management module (33) is further configured to transmit data and transmit it to client devices or other external host devices to the server device (30), such as a memory server.

In the architecture illustrated in FIG. 1c, the server device (30) is a computer on which server software is run. The server software is fragmented into various modules (or software blocks) stored in the memory (32) of the server (30): the communication module (34), the user management module (35), the score management module (36), and the topic management module (37). In one embodiment the modules (34-37) may be configured to communicate with each other via function calls. Preferably, the server (30) is sized to respond to requests from a large number of clients. The queries contain the content that the clients and the server exchange. All or part of the content exchanged between a client and the server can be stored in the memory (32) of the server and / or in the memory (22) of the client.

The server subsystem (30a) comprising the modules (33-37) and the processing unit (31) is operatively coupled to a set of databases: a subject database (38), a database temporary topic data (38a), a user database (39), a temporary comment database (42), and a vote tag database (43). Depending on the embodiment, these databases may be hosted in the memory (32) of the server (30) as shown in Figure 1c, or on a server of remote memory and operatively coupled to the server (30), for example through the input / output interface management module (33).

The communication module (34) ensures the exchange of information between the clients and the server (30). The score management module (36) manages the voting functionality associated with a TrendsTag. The other functions performed by the elements of the server device (30) illustrated in FIG. 1 c are described in detail below.

The processing unit (31) includes one or more processors, which can be any microprocessor, microcontroller, circuit or FPGA (Field Programmable Gate Arrays) component, circuit or ASIC component (Application Specifies Integrated Circuits), signal processing component ( DSP), state machine, or a combination of the previous ones. Depending on the embodiment selected, the server device (30) may include one or more computers, one or more of which may be configured as a multiprocessor computer having multiple processors for performing parallel processing.

Depending on the embodiment, the processing unit (31) may also include, or be able to communicate with, a computer memory, such as, by way of non-limiting example, the memory (32), capable of storing computer program instructions or software instructions which, when executed by a processor, cause that processor to implement the elements described herein.

In addition, the memory (32) can be any type of data storage medium or medium of computer memory coupled to the processing unit (31), such as, for example, a cache memory, a data farm (in English, "data farm"), a data center (in English, "data-center"), a data cloud (in English, "data cloud"), or a combination of the previous ones.

The server device (30) illustrated and described with reference to FIG. 1c is provided by way of example only. Other architectures, operating environments, and configurations are possible. Other embodiments of the server device may include a smaller or larger number of components, and may incorporate all or only some of the features described for the server device (30) of Figure 1c. Therefore, although the processing unit (31), the memory (32), the input / output interface management module (33), the communication module (34), the user management module (35) , the score management module (36), the subject management module (37), the subject database (38), the temporary subject database (38a), the database users (39), the temporary comment database (42), and the vote tag database (43) are illustrated as part of the server device (30), no restriction is placed on the location In particular, in other embodiments, the components 31 - 43 may be integrated into different entities or processing systems.

On the other hand, the architecture of the server device (30) may in some embodiments be distributed, so that the processing unit (31) may be able to control and exchange data with different external components. the entity in which it is incorporated.

A TrendsTag is a computer tag designed to be dedicated to sharing opinion about a topic. Preferably, a TrendsTag represents a subject.

Different types or categories of subject may be associated with a TrendsTag, such as a person, an institution, a community, a place, a product, a project, a company, a work, a show, an event, a social issue etc. A TrendsTag can be associated with any subject, for example: "Windows 10"; "Edward Snowden"; "Abortion"; "Google" "New York," "Gay Marriage," etc.

In a preferred embodiment, the format of a TrendsTag is formed of a character string comprising the subject introduced by a prefix, with no space between the characters of the subject and the characters of the prefix. For example, the prefix can be two exclamation points, with no spaces: "!! ". According to the embodiment, when the subject comprises several words, they can be grouped without space between two words that follow each other, or two words that follow can be separated by a separating character, for example the low-dash character ("_"). The following list provides examples of TrendsTags formed with the prefix "!! And using the underscore character as the separator character: "!! Windows_10"; "Edward_Snowden"; "!! abortion"; "!! Google"; " !!New York " ; "!! Marriage_homosexual", etc.

TrendsTags can be used by any user (and in particular any user), in any language and in any media content. For example, TrendsTags can be used in social media publications, news articles, blogs, forums, opinion sites, and more. The TrendsTags can also be combined with hashtags and / or mentions. The ability of users to share their opinion on a topic through TrendsTags relies on technical solutions that vary according to the type of platforms that users use to communicate. "Platform" means any computer platform (set of equipment on which software is run) comprising an application or set of applications providing users with communication services, such as social networking platforms, platforms e-mail platforms, digital publishing platforms (blogs, online press articles, etc.), online search platforms, Internet browsing platforms, etc. Two types of platforms are distinguished below: those that allow their users to associate a word or a group of words with a URL of their choice, and those that do not allow it.

In a first embodiment, a platform is considered that allows its users to associate a word or a group of words with a URL. In this case, the ability of users to share their opinion by subject through the use of a TrendsTag is based on a resource address associated with the TrendsTag. In the case of users of a platform operating on the Internet, the resource address associated with a TrendsTag may be a URL (Uniform Resource Locator), which points to a web page, which may or may not be accessible to the public. , associated with TrendsTag. The display of this web page on a user device can be configured such that this web page appears on a superimposed display screen of the web page of the platform used which incorporates the TrendsTag.

In one or more embodiments, a Web page associated with the TrendsTag contains credentials on the subject that the TrendsTag represents, in one or more languages depending on the chosen embodiment. The identifying information on the subject that the TrendsTag represents is stored in the subject database (38).

In one or more embodiments, the TrendsTag web page also contains opinion sharing functions. These opinion-sharing functions may for example be provided by one or more opinion collection modules implemented on the server, as illustrated in FIG. 1c by the subjects management module (37) and the management module of the subjects. scores (36) at the server level (30). Depending on the embodiment chosen, the opinion sharing functions may take different forms, and may be for example voting or comment functions as described below. For example, in the case of an opinion-sharing function implemented as a voting function, the user of a TrendsTag accessing the TrendsTag-related web page will be able to cast a vote on the subject. represented by the TrendsTag. Similarly, in the case of an opinion sharing function implemented as a comment function, the user of a TrendsTag accessing the TrendsTag associated web page may provide commentary on the topic. represented by the TrendsTag.

Used in a media content, a TrendsTag and associated with its URL can be identified by the user as an opinion poll systematically formulating the following question: "At this moment, what is your opinion on the subject that the TrendsTag represent ? Depending on the chosen embodiment, this question may not be explicitly formulated to the user. It can indeed be induced by the role and the characteristic form of a TrendsTag associated with its URL. This technical association has the advantage of offering a very condensed opinion poll, facilitating its integration into any type of digital content such as electronic press articles, comments, opinions, descriptions, photos, videos, etc.

Fig. 2 illustrates an example of a TrendsTag web page (40). The "John_Doe" TrendsTag webpage (40) contains information about the TrendsTag, such as a reminder (41a) of the TrendsTag object of the Web page (40), an image (41b), and information complements in text format ("Presidential Candidate, France") (41 c) constituting identifying information of the subject that the TrendsTag represents (in the example the person named John Doe).

The web page (40) illustrated in FIG. 2 additionally includes a voting function, allowing a user to express a polarized opinion on the subject represented by the TrendsTag through a positive vote by clicking on the "UP" button (42a) or a vote negative by clicking on the "DOWN" button (42b). The web page (40) includes score information representing, for example, the percentage of positive votes counted and, in some embodiments, validated.

The web page (40) illustrated in FIG. 2 further comprises a section (43) grouping comments (44a, 45a, 46a), generated by the users and / or, depending on the embodiment, collected in digital contents. In one or more embodiments, the Web page (40) further comprises a voting function for each comment, in the form of buttons "UP" (44b, 45b, 46b) and "DOWN" (44c, 45c, 46c ) associated with each comment (44a, 45a, 46a), and a vote counter for each "UP" button (44b, 45b, 46b) and "DOWN" (44c, 45c, 46c). As illustrated in FIG. 2, comments can be ranked by popularity, based on their number of positive and / or negative votes.

The highest popularity rating (44a) was generated by the pseudo user "@Bowercott" as of 25/03/2016. It is associated with a vote "DOWN" on the TrendsTag "!! John_Doe" that it completes.

The following comment (45a) is a response to the comment (44a) of the user "@Bowercott", generated by a pseudo user "@Supercow".

The third comment (46a) corresponds to a tweet identified by a software engine of the Web Crawler type because it contains the TrendsTag "!! John_Doe". In one or more embodiments, the server (30) can indeed be configured to scan continuously or periodically digital publications such as a tweet to detect whether their content contains a voting tag and a subject tag, or a TrendsTag . In the case where a voting tag and a subject tag or a TrendsTag is detected in a digital publication, the contents of this publication are copied in whole or in part from the memory (32) of the server (30) and then added in all or part as a comment on the TrendsTag web page.

Other comments (not shown in the figure) can be inserted as they are created in the "JohnDoe" TrendsTag Web page (40). Depending on the embodiment, these comments are not necessarily associated with a vote marker. In the embodiments or all or part of the comments are not associated with a vote marker, these comments are not ranked by popularity rank, but can be according to other criteria, such as the date of their creation.

In a second embodiment, a platform is considered that does not allow its users to associate a word or a group of words with a URL. In this case, the ability of users to mark their opinion by subject is based on the TrendsTags without their URL and / or on a method described below with reference to the second embodiment. The method allows a user to polarize their opinion about a subject.

Referring to Figure 1c, in one or more embodiments, a user database (39) groups registered users to use the TrendsTags and / or certain associated services, such as for example the online creation of TrendsTag. The user database (39) may be stored on the server (30), as shown in Figure 1c, or stored on a remote memory server operatively coupled to the server (30). The user management module (35) manages the user database (39). The user management module (35) can be configured to be a brick of the server software.

Each user registered in the user database (39) corresponds in one or more embodiments to a user identifier in the user database (38) (preferably unique), a registration identifier (for example a email address) and a password, which provide registration and authentication functions prior to accessing TrendsTag services. Registration and authentication by e-mail address and password can be replaced by a third-party solution, such as Facebook Connect®. The identifier of a user registered in the user database (39), as well as his e-mail address, are not accessible to other registered users.

In what follows, the term "user" will generally refer to a user of TrendsTag not necessarily registered or registered in the user database (39). One or more of the associated services described below may further be restricted to users enrolled in one or more embodiments, or may be open to all users in other embodiments.

With reference to FIG. 1c, the server subsystem (30a) is operatively coupled to a subject database (38), which may be in some embodiments an integral part of the server device (30).

The subjects management module (37) is a brick of the server software configured inter alia to manage the database of subjects (38), and control one or more software robots (in English, "web crawlers") configured to explore the content of publications on the Internet, and in particular the contents of publications on social networks, as explained below.

In one or more embodiments, the subject database (38) can be powered by integrating third-party subject databases, that is, databases generated by third-party systems into the TrendsTag system ( such as the system (10) of Figure 1 a). Before being injected into the subject database (38), the elements of the third-party subject databases are injected into a temporary subject database (38a). The temporary subject database (38a) may in some embodiments be stored in the memory (32) of the server (30), and the subject management module (37) may be configured to perform the database based processing data temporary topics (38a) relating to the integration of a third database of subjects. Processing of a temporary subject database (38a) by the subject management module (37) may include normalization and cleaning of topics in the temporary subject database (38a). The normalization operation includes formatting the new subject so that it can be compared with topics recorded in the subject database (38). Curation aims at avoiding the introduction of duplicates in the database of subjects (38). Depending on the embodiment, the multilingual translation of subjects of a temporary subject database (38a) as part of its integration into the subject database (38) can be done manually and / or automatically . Once these operations have been performed (normalization, cleaning, and translation into several languages), the subjects management module (37) assigns an identifier, which may be unique in one or more embodiments, to each subject and then integrates the subject into the subject database (38).

In one or more embodiments, a "subject" object of the subject database (38) includes a "name" field filled with a character or a string of characters whose semantic content contains the only information necessary for the identification of the subject by a human being. The same "subject" object can be associated in the subject database (38) with several "name" fields, for example to associate a subject with his name written using different alphabets. In one or more embodiments, the same subject object may be associated in the subject database (38) with synonyms.

In one or more embodiments, the subject management module (37) is configured to generate a TrendsTag from the content of at least one "name" field of each "subject" object listed in the subject database. (38). In one embodiment, a TrendsTag is generated for each "name" field of each "subject" object of the subject database (38).

In one or more embodiments, the generation of a TrendsTag from the content of a "name" field can be performed by preceeding the content of the "name" field considered by the prefix chosen for the TrendsTags (for example, two exclamation points, no space: "!!"). In order to comply with a format defined for the TrendsTags (format that can be defined for several languages, by language and / or by country), it is possible in some embodiments, in the case where the content of the "name" field is a group of words, replace the spaces separating the words by predefined separation characters, for example "low dash" characters, to form a tag consisting of a set of characters without spaces between them.

The generated TrendsTags may be logged, in one or more embodiments, in the subject database (38), within data structures associated or not with the subject data structures in the subject database (38). ), depending on the embodiment.

In one or more embodiments, the subject management module (37) is configured to generate a URL for one or more or even for each TrendsTag recorded in the subject database (38). The generated URL for a TrendsTag may contain in some embodiments the identifier of the subject to which the TrendsTag is associated.

In one or more embodiments, a user can access in his language information related to a TrendsTag through a client (20) having an application (28) capable of accessing a network resource (typically an Internet browser application that can access a web page from the URL of the page) to which the TrendsTag associated URL points.

The request sent by the client (20) to the server (30) preferably contains the language of the client (20) of the user. The subject management module (37) of the server (30) can be configured to process the request on receipt thereof by the server (30), and send to the client (20) of the user at the origin of the request a response containing information attached to the TrendsTag, possibly accompanied by a cookie. In the embodiment where the request sent by the client (20) to the server (30) contains the language of the client (20) of the user, the information sent in response are sent in the language indicated in the request.

The subject management module (37) can be configured to manage the subject database information (38) that can be transmitted to a user in response to a request. When the Web page associated with a TrendsTag is public, the TrendsTag's public URL will return to its public web page and in the user's language. In one or more embodiments, the application (28) of the customer (20) of the user can be configured to transmit, on click or pointing of the user on the TrendsTag, a request addressed to the URL of the TrendsTag on which he clicked or pointed. For example, the user may have clicked the URL associated with a TrendsTag that was inserted in a digital news article or from the results of a search engine that indexed the URL and all or part of the page public web of TrendsTag.

Figure 3a shows an example of a Web page (50) whose content incorporates the TrendsTag "!! John_Doe". Figure 3b shows an example of "John_Doe" TrendsTag information superimposed on the web page of FIG. 3a when the user clicks or points to the TrendsTag "!! John_Doe".

In the example illustrated in Figure 3b, the information relating to the TrendsTag is displayed as a window (51) containing a reminder of the TrendsTag (51 a) object of this information ("!! John_Doe"), an image (51 b), and additional information in text format ("Presidential Candidate, France") (51 c).

The information on the TrendsTag "!! John_Doe" displayed in the window (51) can be, in one or more embodiments, extracted from the information recorded in the subject database (38) for this TrendsTag, for example in connection with with the webpage of this TrendsTag stored in the topics database (38). The superimposed window (51) can thus be generated as an extract from the TrendsTag web page containing a predetermined subset of elements of this web page.

In one or more embodiments, the generation of the web page (50) must include the association of each TrendsTag inserted into the content of the web page (50) with its URL. Software instructions, for example Javascript code, are also integrated into the script of the web page (50) when it is written. These software instructions are configured to execute each time the web page (50) is displayed by an application (for example a browser), and, when executed, to identify in the web page (50) the URLs of TrendsTag embedded therein, and generate for each identified TrendsTag URL a request to the TrendsTag system server (such as the server (30) of Fig. 1c) to retrieve information about the TrendsTag, including information to be displayed in the overprint window (51).

In one or more embodiments, the content of the public web page of a TrendsTag is open to automatic programs for scanning web page contents (in English, "Web crawlers"), also referred to as "indexing robots" or "bots". indexers ". Search engines use Web crawlers to index the content of web pages. Web crawlers browse websites regularly and automatically to discover new URLs. They follow the hypertext links that link the web pages to each other, one after the other. Each identified web page is indexed in the search engine database. The public web pages of the TrendsTags can in particular be indexed and their content be present in the results of the search engines.

In order to improve the identification of a TrendsTag by a human being, each subject of the database may be modified or supplemented in one or more embodiments by additional information (text, photos, videos, sounds, etc.). ) by the Internet users. The additional information is transmitted by the Internet users from the TrendsTag web page and addressed to the topics management module (37) by means of a request from the customer (20) of the user. In one or more embodiments, the subject management module (37) may be configured to store the information thus transmitted in the temporary subject database (38a). The subject management module (37) may further be configured to normalize and clean this information. During curation, if for the same subject the subject management module (37) does not reveal any duplicate information with the subject database (38), the additional information is associated with the subject in the database of subjects (38). subjects (38), and translated into embodiments manually and / or automatically.

In one or more embodiments, a user can create a TrendsTag through a client (20) and a (typically public) URL for creating TrendsTag. The TrendsTag creative URL points to a web page that contains an online TrendsTag creative form in which the user populates the subject's name in a form field in their own language. In one embodiment, the user is prompted by the TrendsTag creation form to provide any other additional information (text, photos, videos, sounds, etc.) useful for identifying the TrendsTag to be created. Once the request for creating TrendsTag submitted by the user's client, it is transmitted to the server where it is processed by the topics management module (37).

Figure 4 shows an example of a form (60) for creating a TrendsTag online. The form contains a "Subject Name" field (61a), a "More Information" field (61b), a file addition button (61c), and a button (61d) to complete the creation of TrendsTag.

The content of the "Subject Name" field is sent to the temporary subject database (38a), and processed by the topic management module (37). The subject management module (37) normalizes and cures this content. The normalization operation aims at formatting the new subject so that it can be compared with the subjects recorded in the subject database (38). The curation aims for her avoid the introduction of duplicates in the subject database (38). The treatments performed on the new subject therefore include a comparison with the existing one in the subject database (38). In the case where the subject subject of the query for creating TrendsTag does not exist in the subject database (38), it is injected with an identifier, which may be unique in one or more embodiments, in the subject database (38). Otherwise, the query for creating TrendsTag is ignored, the temporary subject database (38a) is cleared of the creation request information, and the user may in one embodiment receive a notification of that request. Failure to create the corresponding TrendsTag for the desired subject, possibly accompanied by or containing information about the reason for the failure (eg information that duplication is not possible).

In one embodiment, when a user's request is rejected by the server (30), the subject management module (37) can be configured to identify in the subject database (38) one or more TrendsTag that correspond to the subject for which the user is looking to create a TrendsTag. The search in the subject database (38) can be performed based on the identifier of the subject concerned, in order to find all the TrendsTags recorded in the subject database (38) for this topic identifier. All or part of the results of this search in the subject database (38) can then be presented to the user initiating the request, for example by displaying on the user's client device.

A user creating a TrendsTag through the public TrendsTag creative URL may decide that the TrendsTag is private. In this case, when the creation of TrendsTag is successful, a data structure corresponding to a TrendsTag is created, but with a URL to which the "Private" attribute is associated. This way, only users who know the TrendsTag's private URL can access the TrendsTag's private web page. The private URL of the TrendsTag thus created and the content of its web page are not open to Web crawlers.

In the first embodiment described above, the web page (public or private) of a TrendsTag can be configured to allow a user who consults to express his opinion on the subject corresponding to the TrendsTag, in the form of a vote for example, or the selection of an opinion among a plurality of predefined opinion and which are presented to him to make a selection from the web page.

In the case of a vote, the web page of a TrendsTag allows a surfer access to one or more voting functions according to the embodiment. For example, a TrendsTag webpage may allow a user to access two voting functions, corresponding respectively to a positive or negative opinion about the topic associated with the TrendsTag.

In one or more embodiments, users vote from a client application, such as the application (28) shown in Figure 1b.

The score management module (36) can be configured to be a brick of the server software. Depending on the embodiment, the score management module (36) may be configured to generate voting URLs and / or expose the voting URLs.

In one embodiment, each voting URL generated by the scoring module (36) includes the subject identifier associated with the TrendsTag and a voting status. For example, a positive voting URL may contain the subject identifier and the positive voting status, and a negative voting URL may contain the subject identifier and the negative vote status.

In one or more embodiments, the score management module (36) can be configured to expose a voting URL for each voting state (here the terms "voting status" and "voting function" are used interchangeably. ") Provided in the voting feature associated with the TrendsTag. For the non-limiting example of bipolar voting, in one embodiment, the score management module (36) exposes a positive voting URL associated with the positive voting function of a TrendsTag and vice versa for the negative voting function. .

The voting function is offered to users via the display of all or part of the web page associated with the relevant TrendsTag, a display that can appear on a screen of the user equipment superimposed on the platform that integrates the TrendsTag. . In one or more embodiments, the Web page associated with the TrendsTag includes one or more hypertext links respectively corresponding to the voting functions proposed to the users for this TrendsTag and associated respectively with the voting URLs.

The score management module (36) can also be configured to store, once a vote has been expressed by a user, in the memory (32) of the server a positive voting state for a positive vote and a negative vote state for a negative vote.

In one embodiment, the voting functionality from the TrendsTag associated web page is configured to allow only users registered in the user database (39) to vote. The management module scores (36) is then configured to verify that a vote expressed for a TrendsTag is from a user registered in the user database (39), before engaging other processing related to the vote.

In one or more embodiments, a click of a user (for example a user) on the voting function of the web page of a TrendsTag is equivalent to clicking on the voting URL associated therewith. The user's vote is transmitted from the user's client to the score management module (36) executed by the server by means of a request, for example a request of type GET or of type POST of the HTTP protocol. (in English, "HyperText Transfer Protocol") or HTPPS (in English, "HyperText Transfer Protocol Secure"). The request may in certain embodiments contain the URL of the click function clicked, a client address (for example the IP address of the client), an identifier of the user (in the non-limiting case of a user registered), and in one or more embodiments a connection indicator. The exchange of connection cookies between the server (30) and the clients of the users not registered in the user database (39) makes it possible to limit by vote the votes and the comments on a TrendsTag associated with a URL or associate a vote and a comment from the same client.

In one or more embodiments, the voting functionality is set and the scoring module (36) is configured to provide that users can vote multiple times on the same TrendsTag. For example, a user is likely to vote more regularly on a politician than on a restaurant. In this case, the score management module (36) is further configured to manage a user authorized polling rate parameter and TrendsTag. This parameter may according to the embodiment set a threshold number of votes for the TrendsTag considered above which a vote of the user will no longer be taken into account. This vote count threshold defines a sliding time window, within which votes are counted and compared against the threshold, and outside of which votes are no longer taken into account for this frequency control function. vote.

The score management module (36) can be configured to, in the event that the user's vote does not fall within the authorized polling frequency, send a response to the user's client warning him of the invalidation (or not taking into account) of his vote.

In one or more embodiments, the score management module (36) is configured to assign to each validated or invalidated vote an identifier of vote, preferably single, generate a "t_vote" vote and vote weight, and recalculate the score of at least one voting state for the subject. For example, in the case of a bipolar vote, the score management module (36) can be configured to recalculate the score of positive votes on the subject. In one embodiment, the value of the "t_vote" parameter is provided by the server clock. When the vote is invalid, the vote does not count in the calculation of the score, and for this purpose the weight attributed to it is zero.

The attribution of an identifier for a invalid vote can advantageously be used to perform statistics on all the votes received and processed by the opinion-sharing system. However, it may not be provided in some embodiments.

In one or more embodiments, the score management module

(36) can be configured to address each vote to the topic management module

(37). The subject management module (37) may be configured to store back in the subject database (38) each vote for the corresponding subject. The data structure of a vote recorded in the topic database (38) for a corresponding topic may include the following: vote identifier, voting status, vote "t_vote" instant, client address, the user (in the non-limiting case of a registered user), the subject identifier, and in some embodiments a cookie to limit the client (20) of a non-registered user, for example, which user, to vote several times from the web page of a TrendsTag.

In one or more embodiments, a user who is not registered in the user database (39) can use a TrendsTag to express his or her opinion on the topic corresponding to the TrendsTag. In this case, the score management module (36) can be configured to associate the vote of the non-registered user with a voting weight lower than the voting weight (s) associated with the votes sent by the registered users. The voting data structure stored in the subject database (38) may be the same as that of a registered user, except for the user identifier, for a vote cast by a registered user.

In one or more embodiments, the TrendsTag webpage includes the score of positive votes for the topic corresponding to the TrendsTag, so that the score is accessible from the web page of a TrendsTag. Preferably, the score of the web page of a TrendsTag is updated with each new vote or vote slice stored in the subject database (38) for the subject corresponding. The score management module (36) can be configured to manage by TrendsTag the number of votes that includes a vote slot and from which a score update calculation is performed by the management module. scores (36).

In the first embodiment described above, the frequency of the votes of the same user can be monitored, so that when the user sends his vote from a client through a request to the management module scores (36) and that his vote enters the authorized polling frequency, the subjects management module (37) can be configured to address a response to the client from which the vote is issued allowing the user to formulate a vote. comment. In this case, the subject management module (37) can be configured to generate comment URLs.

The subject management module (37) may be configured to store a comment identifier in the memory (32) of the server. The URL of a comment can then be composed of the following attributes: comment identifier, vote identifier. The user can include in his comment text and / or media content (photos, videos, sounds, etc.). From the client, the user sends his comment to the topic management module (37) by means of a request. The request may for example include the content of the comment, the URL of the comment, and in some embodiments a cookie to associate the vote of the client (20) of a non-registered user, for example any user, to the comment of the same customer from the web page of a TrendsTag.

Upon receiving the request, the subjects management module (37) assigns an identifier, preferably unique, to the comment and generates an instant "t comment" of the comment. The instant "t comment" can for example be provided by the server clock. The subject management module (37) stores the comments in the subject database (38). The data structure of a comment recorded in the subject database (38) for a corresponding subject may comprise, as appropriate, the following elements: comment identifier, comment content, comment "comment" instant, comment identifier vote, subject identifier, client address, user identifier (in the non-limiting case of a registered user), and login indicator.

In one embodiment, a TrendsTag web page includes comments related to the topic of the TrendsTag, so that each comment is accessible from the web page of a TrendsTag. The web page of a TrendsTag is therefore updated with each new comment stored in the database of subjects (38) for the corresponding subject. The score management module (36) can also be configured to classify the comments according to predefined relevance criteria.

A TrendsTag corresponding to a subject, a comment may in one or more embodiments be associated in the subject database (38) to a TrendsTag and the subject it represents.

In the case where two or more users write the same comment about the same subject and at a very close time interval, their respective comments may be in one embodiment recorded and published on the TrendsTag web page, even though that they are identical.

Knowing that a user's opinion of a subject may oscillate over time (for example from positive to negative), the server software may be configured in one or more embodiments so that the right for a user to write a comment expires after a predefined period. Indeed, the shorter the time between the instant "t_vote" of his vote and the moment when the response containing his comment is addressed to the subjects management module (37) is short, the less the opinion of the user may vary. . The predefined period can for example be set to 24 hours. Once the deadline expires, the user is allowed to vote again on the TrendsTag from the TrendsTag web page.

In one or more embodiments, a user who is not a user registered in the user database (39) may, following his vote, comment on the topic corresponding to the TrendsTag. The topic management module (37) associates the comment with the user's customer vote and the subject database (38). In this case, the data structure of a comment recorded in the subject database (38) is the same as that of a registered user, except for the user identifier.

In one embodiment, the TrendsTag web page includes comments related to the subject of the TrendsTag. The score management module (36) can be configured to downgrade the comments from the Internet users to the benefit of comments from users registered in the user database (39).

As part of the second embodiment described above, it is possible to share his opinion on a subject through a TrendsTag from a social media, without going through a TrendsTag URL, nor be a registered user in the database of users (39). For social media users, including social networks, the use of TrendsTags is particularly useful to mark by subject his opinion or mark by subject a fact that could influence the opinion of the Net surfers. Users of these platforms can indeed integrate in their publication the TrendsTags of their choice. For example, the publication "Extraordinary: JohnDoe, unaware that the microphone is still plugged, humming after announcing his resignation" incorporates the TrendsTag "!! JohnDoe".

The proposed method of processing TrendsTags embedded in publications uses the ability of machines to isolate tags for indexing these publications. The subject management module (37) is thus configured in one or more embodiments to index each publication that is identified as containing one or more TrendsTags. The method of indexing publications may be similar to the indexing of web pages by a search engine. The indexing work can thus be based on an embodiment on "Web crawler" or, if the platforms allow it, by exploiting their API, in particular that of Twitter®, Instagram® and Facebook®.

Each indexed publication is injected into a temporary comment database (42). The temporary comment database (42) is preferably configured to store multilingual comments. In one embodiment, the subject management module (37) is configured to manage the temporary comment database (42). The temporary comment database (42) may further be stored in the server's memory (32) (30), or in a remote server operatively coupled to the server (30).

In one embodiment, each indexed publication is stored with the URL of its source page. In the absence of the TrendsTags URL in publications, the topics management module (37) is configured to associate with each TrendsTag the subject it is supposed to represent in the context given by each publication. This work is particularly applicable to distinguish synonyms and polysemic subjects.

For example, the two following TrendsTag will be identified as synonyms, being related to the same subject: "!! UnitedStates" and "!! USA". In contrast, the TrendsTag "!! Lincoln" will be identified as polysemic, as it may for example represent the film of Steven Spielberg, the state capital of Nebraska or the historical figure Abraham Lincoln.

The process of identifying a subject and its meaning in a text is called recognizing named entities. At every TrendsTag of a publication potentially corresponds to a named entity. In a text, a named entity can be: a person, a company, an organization, a work, a place, etc. Named entity recognition may in one or more embodiments rely on existing technologies, such as Alchemy technology from IBM® or TextRazor®. These technologies use the most advanced natural language processing methods associated with artificial intelligence techniques, in order to: distinguish, analyze, understand and extract the semantic content of textual resources. Both technologies offer acceptable levels of Named Entity recognition in the most widely spoken languages on the internet: English, Spanish, Russian, French, German, Italian, Swedish. These languages represent 40% of Internet users worldwide.

Once identified, the named entities of a TrendsTag integrated into an Internet publication are reconciled by the topic management module (37) of the items recorded in the subject database (38) to determine a match between the TrendsTag and a subject listed in the subject database (38).

In the case where a publication contains several TrendsTags, the subject management module is configured to determine whether a TrendsTag contained in the publication may be associated with a subject listed in the subject database (38), based on the subject that he is supposed to represent in the publication.

Once an association is identified between a TrendsTag detected in a publication and a subject in the subject database (38), the TrendsTag is considered identified.

In one or more embodiments, if a publication contains multiple TrendsTags, it is sufficient that at least one TrendsTag of the publication is identified for the subject management module (37) to associate the publication as a comment of the identified TrendsTag. The subject management module (37) assigns an identifier (preferably unique) to the comment. The subject management module (37) stores the comment in the subject database (38) and as appropriate associates it with the only identified TrendsTag in the publication or with all the identified TrendsTags, for example by adding the comment to the list of comments included in the webpage of each TrendsTag, as illustrated in Figure 2 with comment (46a) from a tweet. The data structure of a comment stored in the subject database (38) may include the following: comment identifier, comment content, each Identified topic identifier of TrendsTag and, depending on the type of comment, a URL from the comment.

Each comment captured in this way is accessible from the web page of each identified TrendsTag. As noted above, the web page of a TrendsTag can be updated with each new comment stored in the subject database (38) for the corresponding topic.

In the second embodiment described above, the opinion-sharing methods proposed herein also allow voting on a topic corresponding to a TrendsTag from a social media without passing through a public URL of the TrendsTag or being a registered user in the user database (39). To do this, users of social media (and especially those of social networks) use the method of publication described below:

This method advantageously makes use of the ability of the TrendsTags to symbolize a sentence without formulating it. In the case of the vote on a TrendsTag, the following statement is made through the use of the TrendsTag: "At this moment, my opinion regarding such or such subject (s) ) is positive (or negative). The polarity of the opinion and the subject are reduced to a label (a "tag") and are therefore technically accountable by a machine. The feeling and the subject being each reduced to a tag, the proposed method allows in this embodiment a user to polarize his opinion of a subject in a sentence that may be as short as possible, reduced to a subject tag and a voting tag.

In one or more embodiments, a publication is identified as being polarized when it includes a polarity marker. The proposed method uses one or more polarity markers, a polarity marker being defined as a written vote tag without space and introduced by a predefined prefix. For example, the vote tag can be a hashtag. A positive vote tag is positive, as shown by the following examples: #UP and #POUR. A negative vote tag is negative, as the following examples show: "#DOWN" and "#CONTRE". The score management module (36) manages a vote tag database (43), preferably multilingual. The vote tag database (43) may further be stored in the memory (32) of the server (30), or in a remote server operatively coupled to the server (30). The score management module (36) can also be configured to give the positive vote tags a positive state and the negative vote tags a negative state.

In a publication, the subject is represented by a tag, which can be a hashtag, a mention or a TrendsTag. The method allows you to associate a voting tag with one or more tag (s) of subjects but does not associate multiple voting tags with the same subject tag.

The method relies on the ability of machines to isolate tags to isolate voting tags and topic tags. Social network users use this method and rely on the machines' ability to isolate voting tags and topic tags to index their posts.

There are two voting channels: one where the subject tag (s) are placed after the vote tag ("vote - subject (s)"), and the other where the tag (s) ( s) of subject are placed before ("subject (s) - vote"). An example of a voting chain where the subject tag (s) are placed after the vote tag is given below:

«#UP @RestauBestSushi for this #SushiBurger discovery»

In this example, the vote tag "#UP" is associated in the publication with the two topic tags (taking for one the form of a mention and for the other the form of a hashtag) "@ Restau BestSushi And #SushiBurger.

These two voting channels can be combined together to give longer chains, for example:

• subject (s) - vote - subject (s) - vote

• subject (s) - vote - vote - subject (s)

• vote - topic (s) - vote - subject (s)

The following is an example of subject (s) - vote - subject (s) - vote:

• #Andromedid takes market share #UP in front of #DOE_OS #DOWN

In this example, the vote tag "#UP" is associated in the publication with the tag of the subject (having the format of a hashtag) "#Andromedid", and the tag of vote "#DOWN" is associated in the publication with the tag of subject (having the format of a hashtag) "#DOE_OS".

The voting channels used can be as long as desired. In the particular case of two contiguous voting chains, the first voting chain being of the "vote - subject (s)" type, and the second voting channel being of the "subject (s) - vote" type (for example the channel following vote: "#UP #SourcePower, #RATP #Greve #DOWN "), so that the publication can be indexed according to the modalities described below, the user can in a preferred embodiment introduce between the two chains of votes a separation marker. Depending on the embodiment, the separation marker may be a hyphen a semicolon the end of a parenthesis and / or the beginning of another parenthesis ") (" or any other separation marker usually used in one language to isolate an element in a sentence The score management module (36) may be configured to manage a separation marker database, preferably multilingual.

The vote tag database (43) may further be stored in the memory (32) of the server (30), or in a remote server operatively coupled to the server (30).

The following examples illustrate user posts that associate a voting tag with a topic tag:

«The # SuperPhoneB9 has an unequaled autonomy #UP»

"#UP #SpringFill #DOWN # Company_Transport_John_Doe #DOWN # Strike".

In the first example, the vote tag "#UP" is associated in the publication with the subject tag (having the format of a hashtag) "# SuperPhoneB9".

In the second example, the vote tag "#UP" is associated in the publication with the tag of the subject (having the format of a hashtag) "#SepeurPompier", the vote tag "#DOWN" is associated in the publication with tag of subject (having the format of a hashtag) "# Company_Transport_John_Doe", and the tag of vote "#DOWN" is associated in the publication with the tag of subject (having the format of a hashtag) "# Strike".

The following examples illustrate user posts that associate a voting tag with multiple topic tags:

«Movies seen this week #UP !! TheMovieOne and !! TheBestMovie»

"#DOWN # Company_Transport_John_Doe # Strike".

In the first example, the "#UP" vote tag is associated in the publication with the two topic tags (taking the form of TrendsTag) "!! TheMovieOne "and" !! TheBestMovie ".

In the second example, the vote tag "#DOWN" is associated in the publication with the two topic tags (in the form of hashtags) "#Transport_John_Doe_Company" and "# Strike".

In one or more embodiments, the subject management module (37) is configured to index each identified publication with one or more voting tags. The indexing work is based on "Web crawler" or if platforms allow it on their API, including that of Twitter®, Instagram® and Facebook®.

Each indexed publication is injected into the temporary comment database (42). The subject management module (37) can be configured to manage the temporary comment database (42). Each publication is stored with the URL of its source page.

Each topic tag in a publication potentially matches a named entity. The named entities of each topic tag are reconciled by the subjects management module to the subject database. If a publication contains multiple topic tags, the topic management module may or may not associate a topic tag with the subject it is supposed to represent in the publication. A subject tag associated with its subject in the subject database means an identified subject.

If a publication contains multiple topic tags, it is sufficient that at least one subject of the publication is identified for the topic management module (37) to associate the publication as a comment of the identified topic. The subject management module (37) assigns an identifier (preferably unique) to the comment. The subject management module (37) stores the comment in the subject database (38). In the subject database (38), the subject management module (37) associates the comment with the only identified subject or with all identified topics of the publication.

As described above for embodiments relating to a comment on a non-URL related TrendsTag, the data structure of a comment stored in the subject database (38) may include the following: comment, content of the comment, each identified TrendsTag topic ID, and the URL of the comment.

Each comment thus captured is accessible from the public web page of each subject TrendsTag identified as illustrated by the comment 46a integrated in the web page 40 of FIG. 2.

The subject management module (37) can also be configured to assign each vote tag an identifier (preferably unique) and a vote status. If a publication contains multiple topic tags for the same voting tag, the topic management module (37) can be configured to associate or not a vote tag to all topic tags. The subject management module (37) stores each vote and its state in the subject database (38). If a vote could not be associated with a topic tag, the vote is not stored in the subject database (38).

The vote data structure stored in the subject database (38) may include the following: vote identifier, each identified topic identifier, comment identifier, and comment source URL.

The subject management module (37) sends to the score management module (36) each vote of the subject database (38). The score management module (36) defines the weight of the vote and recalculates the score of the positive votes for each TrendsTag of each identified subject of each vote.

Figure 5a shows an example of a data structure for a "subject" object stored in the subject database (38). The data structure associates with each object "subject" a field containing an identifier, which is preferably chosen to be unique, one or more language fields, for each language field one or more country fields, for each country field one or several name fields, and for each name field a TrendsTag field and one or more additional information fields which may each contain media content (text, image, video, etc.).

Figure 5b shows an example of a data structure for a "TrendsTag" object stored in the subject database (38). The data structure associates with each "TrendsTag" object a field containing a name corresponding to the subject to which the TrendsTag is associated, a field containing a URL, which itself contains an identifier of the subject to which the TrendsTag is associated, and a field of information. attribute whose content indicates whether the URL is public or private. A field containing a TrendsTag-related web page can be included in the data structure of a TrendsTag. This data structure further comprises a vote field (which corresponds to a vote identifier (not shown in the figure)), which comprises one or more voting states which are respectively associated with voting URLs. The data structure of a TrendsTag illustrated in Figure 5b further comprises a comments field, which includes one or more fields containing a comment (each identified by a comment identifier (not shown in the figure)), which can be either no, depending on the case, associated with a voting URL field corresponding to a voting state, and / or associated with a URL from the comment. Embodiments of the proposed methods may be, at least in part, implemented on virtually any type of computer, regardless of the platform used.

Although the functionality of the client and the server, and in particular the different modules of the elementary parity node server, and in particular the output message generator, are described in the form of software, they can be implemented under material form or in the form of a combination of hardware and software instructions.

Depending on the embodiment chosen, certain acts, actions, events or functions of each of the methods described in this document may be performed or occur in a different order from that in which they were described, or may be added, merged or not to be performed or not to occur, as the case may be. In addition, in some embodiments, certain acts, actions or events are performed or occur concurrently and not successively.

Although described through a number of detailed exemplary embodiments, the proposed opinion-sharing method, and the equipment and software for implementing the method include various alternatives, modifications, and enhancements that will be apparent to the user. those skilled in the art, it being understood that these various variants, modifications and improvements are within the scope of the invention.

In addition, although the various embodiments described above implement the use of the Internet network by Internet users, the invention is also applicable to any network on which users can share resources such as media content.

In addition, various aspects and features described above may be implemented together, or separately, or substituted for each other, and all of the various combinations and sub-combinations of aspects and features are within the scope of the invention. 'invention. In addition, some of the systems and equipment described above may not incorporate all of the modules and features described for the preferred embodiments.

Claims

1. An opinion sharing method implemented in a communication network comprising a server capable of communicating with one or more client devices, and comprising:

updating the opinion data associated with said first subject based on the vote or comment received.

The method of claim 1, wherein the first subject is further associated with a first web page and a first resource address in the database;

upon receipt of a request by a client device including the first resource address, the method includes sending the first web page to the client device.

The method of claim 2, wherein the first resource address is inserted into a web page of the service platform.

4. The method of claim 2 or 3, wherein the first web page comprises at least one voting function, and wherein the vote is received by the server via the vote function of the first web page and in which the day of opinion data associated with the first subject includes recalculating a score of at least one voting state for the first subject.

The method of claim 4, wherein a positive voting URL and a negative voting URL are exposed on the first web page, the positive voting URL including an identifier of the first subject and a positive voting status and the Negative vote URL including the identifier of the first subject and a negative vote status.

The method of claim 5, further comprising, upon receipt of a request including a positive polling state or a negative polling state:

generate a comment URL including a comment identifier and a vote identifier;

receiving a request from the terminal comprising the comment formulated and the comment URL for storage in association with the first subject in the database.

The method of claim 1, wherein the first computer tag comprises a subject tag followed or preceded by a vote tag, wherein the subject tag comprises a prefix and a string of characters relating to the first subject and which the vote tag comprises a prefix and a string of characters relating to a vote state, and wherein the vote or comment corresponding to the first subject is the voting state of the vote tag.

The method of claim 7, wherein the first subject is identified from the character sequence by recognizing named entities.

9. Method according to one of the preceding claims, wherein the server database is powered by third-party databases and wherein the storage step comprises receiving at least one subject, formatting the new subject in order to compare it with other topics in the server database, and curation of the new topic so as to avoid any introduction of duplicate topics into the server database.

The method of one of the preceding claims, wherein the storing in the server database comprises receiving a request to create a new subject from a client device;

generation by the server of a new computer tag in association with the new subject, for storing the new subject in association with the new computer tag in the database.

1 1. Method according to one of the preceding claims, wherein the method comprises scanning web pages to identify the first tag, and identifying the first tag in a content of a given web page, updating the data. of opinion includes adding the content as a comment associated with the first topic.

12. Computer program support storing instructions which, when implemented by a processor, lead this processor to implement the steps of the method according to one of claims 1 to 1 1.

13. Server capable of communicating with one or more client devices and with a service platform via a communication network, the server comprising:

at least one database storing subjects, each subject being associated with at least one computer tag and opinion data in the database;

a communication module for receiving a vote or comment relating to a first subject identified on the service platform used by client devices for communicating, the first subject being associated with a first computer tag in the database;

14. A communication system for the sharing of opinion comprising a server according to claim 13 and a service platform capable of communicating with said server and via which client devices are able to communicate.