FR3139211A1 - Method and device for rendering digital content - Google Patents

Abstract

The present invention relates to a method and a device for rendering a set of digital content. For this purpose, first data representative of search criteria are received (41). Second data representative of digital content sources is generated (42) as a result of an automatic search based on the first data. Third data representative of the set of digital contents are obtained (43) by extraction of the sources. Text data associated with the third data is processed (44) using a natural language processing method. The processed text data is classified (45) according to at least one machine learning method. The set of digital content is rendered (46) based on a classification result. Figure for abstract: Figure 4

Description

Method and device for rendering digital content

The present invention relates to methods and devices for rendering digital content. The present invention also relates to a method and a device for searching one or more digital contents. The present invention also relates to a method and a device for classifying data representative of a set of digital content.

Technology background

With the development of the Internet, the volume of information available on a multitude of subjects is extremely large, this information being provided in the form of a multitude of available data sources, in particular through websites (also called websites ), databases of videos, images, text or other digital data content.

To facilitate the search for information on a particular subject, it is known to use one or more search engines powered by one or more keywords entered by the person doing the research.

Searching for information on a particular subject can help in decision-making, for example when deciding on a direction to take in the design and development of a product, for example a vehicle.

However, the sole use of a search engine in the context of research on a particular subject returns a very large amount of data which is sometimes difficult to use for the person who carried out the research due to the volume of data available.

Thus, access to the most relevant information is not guaranteed and decisions based on the results of the research may turn out to be bad.

Summary of the present invention

An object of the present invention is to solve at least one of the problems of the technological background described above.

Another object of the present invention is to improve the search for data on internet and/or extranet type networks.

According to a first aspect, the present invention relates to a method for rendering a set of digital content, the method being implemented by at least one processor and comprising the following steps:

- reception of first data representative of a set of search criteria;

- generation of second data representative of a set of addresses of digital content sources as the result of an automatic search via at least one search engine powered by a keyword search query automatically generated from the first data;

- generation of third data representative of all digital content by a data extraction technique, called “web scraping”, contained in digital content sources;

- processing of text data associated with the third data by automatic natural language processing method;

- classification of processed text data using at least one machine learning method;

- rendering of the set of digital content based on a classification result.

The classification of data extracted from several data sources and processed by an automatic natural language processing method makes it possible to analyze on the same basis a large number of data from different sources to highlight the most relevant to given search criteria.

This makes it possible to improve decision support on a given subject or problem by highlighting the content(s) that best meet the search criteria.

According to a variant, the rendering comprises a display of the set of digital contents on a display device, the display being organized according to the result of the classification.

According to another variant, the set of digital content includes:

- at least one textual content; and or

- at least one video content; and or

- at least one image.

According to an additional variant, the automatic search is also a function of representative evaluation data associated with the digital contents and/or the third data are ordered according to the representative evaluation data associated with the digital contents.

According to yet another variant, the method further comprises a step of automatically translating the text data into a target language determined prior to processing by an automatic natural language processing method.

According to an additional variant, the classification comprises the generation of a score associated with each digital content, the score also being a function of the representative evaluation data associated with the digital content.

According to another variant, the method further comprises a step of receiving representative rendering evaluation data, the representative rendering evaluation data being taken into account in a learning of parameters of at least one classification model implemented. implemented by at least one machine learning method.

According to an additional variant, the at least one machine learning method is of the supervised, unsupervised or semi-supervised type.

According to a second aspect, the present invention relates to a device for rendering a set of digital content, the device comprising a memory associated with a processor configured for implementing the steps of the method according to the first aspect of the present invention.

According to a third aspect, the present invention relates to a computer program which comprises instructions adapted for the execution of the steps of the method according to the first aspect of the present invention, in particular when the computer program is executed by at least one processor.

Such a computer program can use any programming language, and be in the form of a source code, an object code, or an intermediate code between a source code and an object code, such as in partially compiled form, or in any other desirable form.

According to a fourth aspect, the present invention relates to a computer-readable recording medium on which is recorded a computer program comprising instructions for executing the steps of the method according to the first aspect of the present invention.

On the one hand, the recording medium can be any entity or device capable of storing the program. For example, the medium may comprise a storage means, such as a ROM memory, a CD-ROM or a ROM memory of the microelectronic circuit type, or even a magnetic recording means or a hard disk.

On the other hand, this recording medium can also be a transmissible medium such as an electrical or optical signal, such a signal being able to be conveyed via an electrical or optical cable, by conventional or terrestrial radio or by self-directed laser beam or by other ways. The computer program according to the present invention can in particular be downloaded onto an Internet type network.

Alternatively, the recording medium may be an integrated circuit in which the computer program is incorporated, the integrated circuit being adapted to execute or to be used in executing the method in question.

Brief description of the figures

Other characteristics and advantages of the present invention will emerge from the description of the particular and non-limiting examples of embodiment of the present invention below, with reference to the appended Figures 1 to 4, in which:

schematically illustrates a data communication network, according to a particular and non-limiting embodiment of the present invention;

schematically illustrates a process for rendering digital content obtained during a data search via the network of the , according to a particular and non-limiting embodiment of the present invention;

schematically illustrates a network device of the configured for rendering digital content, according to a particular and non-limiting embodiment of the present invention;

illustrates a flowchart of the different stages of a process for rendering digital content implemented by the device of the , according to a particular and non-limiting embodiment of the present invention.

Description of the implementation examples

A method and a device for rendering digital content will now be described in what follows with reference jointly to Figures 1 to 4. The same elements are identified with the same reference signs throughout the description which follows.

There schematically illustrates a data communication system 1, according to a particular and non-limiting embodiment of the present invention.

The system 1 comprises for example a data processing device 10 of the computer or laptop type. According to one variant, the device 10 corresponds to a mobile communication device, for example a smart phone (from the English “Smartphone”) or a tablet.

The device 10 is connected by wired and/or wireless communication to one or more remote devices 101, such remote devices hosting digital content data, for example text data, video content data, image data , etc.

The communication infrastructure connecting the device 10 to the remote device(s) 101 corresponds to one of the following infrastructures or to an infrastructure combining several of the following infrastructures:

- infrastructure of a wireless communication network, for example a 4G or 5G type wireless cellular network or a WLAN type network (from the English “Wireless Local Area Network” or in French “Réseau local sans fil”) ;

- infrastructure of a wired communication network of the WAN type (from the English “Wide Area Network” or in French “Réseau extendée”) or LAN (from the English “Local Area Network” or in French “Réseau local”) ;

- infrastructure of a fiber optic network.

The remote device(s) correspond for example to web servers hosting web resources (web pages, images, videos, etc.). The remote device(s) 101 belong for example to a public network 100 (internet) and/or to a private network (intranet).

There schematically illustrates a process for rendering digital content obtained by analysis and processing of the results of an automatic content search in a set of digital content data sources, according to a particular and non-limiting embodiment of the present invention.

The process of is for example implemented in the system 1, and more specifically in the data processing device 10, for example by one or more processors of this device 10. An example of hardware embodiment of such a device is described in relation to there below.

In a first operation 201, first data representative of a set of criteria for research to be carried out on a particular subject or a specific problem are received.

These criteria are for example filled in by a user via a form generated by the device 10 and rendered, for example displayed, on a screen of the device 10 via a graphical man-machine interface (HMI).

For this purpose, the user enters for example a set of fields using ad hoc interface(s) (for example a keyboard and/or a mouse and/or a touch interface) connected to the device 10.

The field(s) of the form correspond for example to one or more text areas to be filled in (for example with one or more key words), to one or more check boxes, to a selection in one or more drop-down menu lists.

Such a form thus allows the user to provide the device 10 with the elements necessary to carry out the search for digital content on the subject described by the user via the form (keywords, language(s) of the sources to search, dates of publication of sources, etc.).

The form corresponds for example to one or more pages of a HMI of an application executed by the device 10 or to one or more web pages displayed and accessible via a browser executed by the device 10.

Once the form has been completed, the search for digital content is launched, for example following a user action such as a click or a press on a virtual button on the HMI.

The subject of the research corresponds, for example, to automobiles, building construction, health or any other subject.

As an illustrative example, the subject of the research corresponds for example to the field of electric vehicles or even to the field of autonomous vehicles.

In a second operation 202, an automatic search via one or more search engines (for example Google®, Qwant®, Bing®) is launched on the basis of keywords and other criteria of a search query obtained or automatically generated at the first operation 201 via the information provided by the user.

Among the search criteria, a limit on the number of web pages or digital content resources is for example set (for example in the search criteria or as a system default setting).

The search via the search engine(s) is carried out on the internet and/or on a private network such as an intranet.

In a third operation 203, a set of resources or sources of digital content (web pages, video content on platforms such as Youtube® for example, images, advertisements, audio content, etc. and any comments associated with this content) are identified according to any method known to those skilled in the art.

For example, the technique implemented for identifying sources (also called resources) of digital content corresponds to the so-called “Web Scraping” method (or in French “website content extraction method”, also called “harvesting” in English).

Such a method is implemented by executing one or more scripts (for example in python) or programs.

Source identification returns, for example, a set of site addresses hosting one or more digital content (text, video, image, audio, etc.), for example in the form of web addresses or hyperlinks.

According to an optional embodiment, the identification of sources or resources takes into account data representative of the evaluation of digital content associated with the digital content. An evaluation of content corresponds for example to any comment, any assessment (such as “like” (or “aime” in French) or “dislike” (or “does not like” in French), number of stars marking the level of appreciation) made by a person who has accessed and/or consumed the digital content, depending on the availability of these evaluations at the source of the digital content.

In a fourth operation 204, one or more data frames are generated in the form of second data representative of digital content addresses.

The identification of the sources implemented in the third operation returns or generates for example all or part of the following second data:

- a first set of addresses pointing to textual content; and or

- a second set of addresses pointing to video content; and or

- a third set of addresses pointing to image content; and or

- a fourth set of addresses pointing to other types of content (audio, advertising, etc.).

The second representative address data are for example stored in one or more registers of a memory of the device 10, for example transiently or temporarily.

In a fifth operation 205, the content sources identified and generated in operations 203 and 204 are analyzed according to the keywords used for the search, for example according to the “Web Scraping” method.

Textual content (web pages, comments) is searched to identify content including the keywords used in the search.

The images or videos are analyzed (for example by applying automatic learning or machine learning or deep learning methods) to identify objects in images or videos related to the searched keywords.

For example, when the search concerns an electric vehicle, a search for a vehicle type object is carried out in the images and videos to identify content including such an object.

In a sixth operation 206, one or more sets of third data representative of the digital contents identified in the fifth operation 205 are generated and stored in the memory of the device 10.

The third data includes for example one or more of the following sets of data:

- a first set of data representative of the texts of the web pages and/or websites identified as a relevant data source, this first set of data forming a database called a knowledge base; and or

- a second set of data representative of links pointing to the video data stored in memory of the video content identified as relevant; and or

- a third set of data representative of links pointing to the image data stored in memory of the images identified as relevant; and or

- a fourth set of text data (comments for example) associated with video content identified as relevant; and or

- a fifth set of text data (comments for example) associated with images identified as relevant; and or

- a sixth set of data representative of links pointing to the audio data stored in memory of the audio content identified as relevant.

According to an optional embodiment, the representative evaluation data (for example number of “likes” (or “aimes” in French), number of “dislikes” (or “does not like” in French), number of stars) are taken into account to order digital content, for example by organizing the most popular content at the top of the list, the order in the list of content being a function of the evaluation, that is to say that the The most popular content (i.e. the one with the most “likes” and/or the most stars and/or the fewest stars) is at the top of the list while the least popular content is at the bottom of the list.

In a seventh operation 207, the sets of third data are stored in a specific area of a memory of the device 10. For example a memory area is provided for text type contents, another memory area is provided for text type contents video, another memory area is provided for image type content, etc.

In an eighth operation 208, the texts (web pages, comments) extracted from the different content sources identified as relevant are automatically translated by an automatic translation engine from their publication language to a determined target language, the determined target language corresponding for example to English.

According to one variant, a tool for filtering advertisements associated with the content is implemented to remove advertisements from the content to be translated.

The translated texts are for example saved in memory while retaining the text in the original language, the translated text being for example matched with the original text in a correspondence table.

In a ninth operation 209, the translated texts are analyzed and processed using any automatic natural language processing method, called NLP (from English “Natural Language Processing”), known to those skilled in the art.

Such processing is for example implemented according to an automatic learning method, for example implemented via a neural network.

An example of automatic natural language processing is for example described in “Automatic natural language processing and interpretation: contribution to the development of a computer model of interpretative semantics”, published by Ludovic Tanguy (Computer science and language, University of Rennes 1) in 1997.

Text processing using an NLP method aims to prepare and homogenize texts, for example for subsequent processing. NLP processing, for example, identifies elements of the text corresponding to verbs, adjectives and terms that appear most frequently. NLP processing includes, for example, stemming or stemming operations, lemmatization operations, deletion of stop words (for example the words of, of, a, etc.), deletion of punctuation marks, etc.

At the output of the NLP processing model(s), texts prepared or cleaned for the following operations are obtained, such texts being stored in memory (for example temporarily) and associated with the translated texts provided as input to the NLP model(s).

In a tenth operation 210, the text data processed by the NLP model(s) are provided as input to one or more classification models of an automatic learning method. This or these classification models are for example implemented by a neural network.

For example, a supervised classification model is used to classify the data when the latter are representative of a known case or learned from training data prior to the tenth operation 210.

Thus, a supervised type model is based on data processed according to the NLP method as input, on data learned during a prior learning phase and optionally on data representative of the reliability of the source determined according to the source evaluation data.

An unsupervised classification model is used to classify the data when it is representative of a case not encountered during the training phase of the supervised model.

Thus, an unsupervised type model is based on data processed according to the NLP method as input and optionally on data representative of the reliability of the source determined according to the source evaluation data.

Finally, according to another example, a semi-supervised classification model is used to classify the data, whether the latter corresponds to a case encountered in the learning phase or not, the semi-supervised type model being refined at as different data obtained during different researches in the mode of production are classified.

Thus, a semi-supervised type model relies on data processed according to the NLP method as input, on data learned during a prior learning phase on a given subject (or not) and optionally on representative data of the reliability of the source determined based on the source evaluation data.

In an eleventh operation 211, a score (or a probability) is determined for each text provided as input to the classification model(s), the score then being associated with said text.

The score is for example obtained from:

- the number of positive opinions (detected from comments associated with the text) associated with the text;

- the number of positive evaluations (for example the number of “likes”); And

- the probability of belonging to the class obtained at the classification output (for example according to the score function of the so-called Random Forest method (or “Random Forest” in French).

In a twelfth operation 212, the original digital contents associated with the classified texts are recovered from the memory of the device 10.

According to a variant, only the contents having obtained a score greater than a determined threshold (for example greater than 0.6, 0.7 or 0.8) are selected and retrieved from the memory.

In a thirteenth operation 213, a rendering of the digital content obtained in the twelfth operation 212 is implemented for restitution to the user who launched the search.

The rendering includes for example a display of the digital contents selected according to an ordering which depends on the score obtained in the twelfth operation, the digital content(s) having the best score (that is to say best corresponding to the search criteria being placed forward (for example displayed with a larger size and/or at the top of the page (or in the center of the page)).

When the rendering corresponds to a display, the contents are organized on one or more pages according to their respective scores for display on the display screen of the device 10 for example.

In a fourteenth operation 214, a query for evaluating the search result is generated and rendered (for example displayed) to the user who launched the search.

Such a request takes for example the form of a form to be filled in (for example with check boxes) to check whether or not the search result meets the user's expectations.

In a fifteenth step 215, the user's responses to the form are received and stored in the memory of the device 10.

In the absence of a response from the user, it is for example considered that the summary of the search results complies with the user's expectations.

In a sixteenth step 216, the data obtained in response to the form are used to refine the parameters of the classification model(s) implemented in the tenth operation 210.

Such a process thus makes it possible to take into consideration a larger number of sources than in traditional internet search methods, the data obtained from the search being analyzed and processed automatically using machine learning techniques (from English " machine learning") to filter and select the search results considered to be the most relevant.

A summary of the content found is generated taking into account a score, which makes it easier for the user to read the results by emphasizing the content judged to be the most relevant to the search criteria. .

The search result is thus more reliable, even on complex research topics.

There schematically illustrates a device 3 configured for searching for digital content and for rendering a summary of the search result, according to a particular and non-limiting embodiment of the present invention. The device 3 corresponds for example to the data processing device 10 of the .

The device 3 is for example configured for the implementation of the operations described with regard to Figures 1 and 2 and/or the steps of the method described with regard to the . Examples of such a device 3 include, but are not limited to, a computer, a laptop, a smartphone, a tablet. The elements of device 3, individually or in combination, can be integrated into a single integrated circuit, into several integrated circuits, and/or into discrete components. The device 3 can be produced in the form of electronic circuits or software (or computer) modules or even a combination of electronic circuits and software modules.

The device 3 comprises one (or more) processor(s) 30 configured to execute instructions for carrying out the steps of the method and/or for executing the instructions of the software(s) embedded in the device 3. The processor 30 may include integrated memory, an input/output interface, and various circuits known to those skilled in the art. The device 3 further comprises at least one memory 31 corresponding for example to a volatile and/or non-volatile memory and/or comprises a memory storage device which may comprise volatile and/or non-volatile memory, such as EEPROM, ROM , PROM, RAM, DRAM, SRAM, flash, magnetic or optical disk.

The computer code of the embedded software(s) comprising the instructions to be loaded and executed by the processor is for example stored on memory 31.

According to a particular and non-limiting embodiment, the device 3 comprises a block 32 of interface elements for communicating with external devices, for example one or more remote servers. The interface elements of block 32 include one or more of the following interfaces:

- RF radio frequency interface, for example of the Wi-Fi® type (according to IEEE 802.11), for example in the 2.4 or 5 GHz frequency bands, or of the Bluetooth® type (according to IEEE 802.15.1), in the band frequency at 2.4 GHz, or Sigfox type using UBN radio technology (from English Ultra Narrow Band, in French ultra narrow band), or LoRa in the 868 MHz frequency band, LTE (from English “ Long-Term Evolution” or in French “Long-Term Evolution”), LTE-Advanced (or in French LTE-advanced);

- USB interface (from the English “Universal Serial Bus” or “Bus Universel en Série” in French);

- HDMI interface (from the English “High Definition Multimedia Interface”, or “Interface Multimedia Haute Definition” in French).

According to another particular and non-limiting example of embodiment, the device 3 comprises a communication interface 33 which makes it possible to establish communication with other devices (such as other computers or servers) via a communication channel 330. The communication interface 33 corresponds for example to a transmitter configured to transmit and receive information and/or data via the communication channel 330. The communication interface 33 corresponds for example to a wired network of the Ethernet type (standardized by the ISO/IEC 802-3 standard).

According to a particular and non-limiting embodiment, the device 3 can provide output signals to one or more external devices, such as a display screen 340, touch or not, one or more speakers 350 and/or other peripherals 360 (for example a haptic actuator) via output interfaces 34, 35 and 36 respectively. According to a variant, one or the other of the external devices is integrated into the device 3.

There illustrates a flowchart of the different stages of a process for rendering digital content, according to a particular and non-limiting exemplary embodiment of the present invention. The method is for example implemented by a processing device 10 of a communication system or by the device 3 of the .

In a first step 41, first data representative of a set of search criteria are received.

In a second step 42, second data representative of a set of addresses of digital content sources are generated as the result of an automatic search via at least one search engine powered by a keyword search query automatically generated at from the first data.

In a third step 43, third data representative of the set of digital content are generated by a data extraction technique, called “web scraping”, contained in the digital content sources.

In a fourth step 44, text data associated with the third data are processed by natural language processing method.

In a fifth step 45, the processed text data are classified according to at least one machine learning method.

In a sixth step 46, the set of digital content is rendered according to a classification result.

According to a variant, the variants and examples of the operations described in relation to the or 2 apply to the steps of the process of .

Of course, the present invention is not limited to the exemplary embodiments described above but extends to a method of classifying text data associated with one or more digital contents which would include secondary steps without thereby departing from the scope of the present invention. The same would apply to a device configured to implement such a process.

Claims (10)

Method for rendering a set of digital content, said method being implemented by at least one processor and comprising the following steps:
- reception (41) of first data representative of a set of search criteria;
- generation (42) of second data representative of a set of addresses of digital content sources as the result of an automatic search via at least one search engine powered by a keyword search query automatically generated from said first data data ;
- generation (43) of third data representative of said set of digital content by a data extraction technique, called "web scraping", contained in said sources of digital content;
- processing (44) of text data associated with said third data by automatic natural language processing method;
- classification (45) of said text data processed according to at least one automatic learning method;
- rendering (46) of said set of digital contents according to a result of said classification. A method according to claim 1, wherein said rendering comprises displaying said set of digital contents on a display device, said display being organized according to said result of said classification (45). Method according to claim 1 or 2, for which said set of digital contents comprises:
- at least one textual content; and or
- at least one video content; and or
- at least one image. Method according to one of claims 1 to 3, for which said automatic search is further a function of representative evaluation data associated with said digital contents and/or said third data is ordered according to said representative evaluation data associated with said digital contents . Method according to one of claims 1 to 4, further comprising a step of automatic translation (208) of said text data into a target language determined prior to said processing (44) by automatic natural language processing method. A method according to claim 4, wherein said classification comprises generating (211) a score associated with each digital content, said score further being a function of said representative evaluation data associated with said digital content. Method according to one of claims 1 to 6, further comprising a step of receiving (215) data representative of evaluation of said rendering, said data representative of evaluation of said rendering being taken into account in a learning of parameters of at at least one classification model implemented by said at least one machine learning method. Method according to one of claims 1 to 7, for which said at least one automatic learning method is of the supervised, unsupervised or semi-supervised type. Computer program comprising instructions for implementing the method according to any one of the preceding claims, when these instructions are executed by a processor. Device (3) for rendering a set of digital contents, said device (3) comprising a memory (31) associated with at least one processor (30) configured for implementing the steps of the method according to any one of the claims 1 to 8.