Method and system for producing micro-video content dedicated to foreign language learning, and learning method implementing this video content

Abstract

A method (10) for producing micro-video content (7; 7.1, 7.2, 7.3) for language learning, from long-form videos extracted from one or more video content sources (11), implemented by computer, this method comprising the following steps: a search and identification (12), from said video content source(s) (11), of a plurality of video content containing the same predetermined target word or the same predetermined target phrase belonging to said language; an extraction, within each of said video content thus identified, of a video segment (4) containing said predetermined target word or predetermined target phrase; an enrichment of each video segment thus extracted, by adding metadata and a contextualized test, in relation to said predetermined target word or predetermined target phrase, said enriched video segment constituting a micro-video learning content. See Figure 2

Description

Description Title of the invention: Method and system for producing micro-video content dedicated to learning a foreign language, and learning process implementing this video content FIELD OF INVENTION

[0001] = The present invention relates to a method and system for producing micro- video content dedicated to learning a foreign language. It also aims to a foreign language learning system implementing this process, thus than a language learning process implementing these micro-video contents. STATE OF THE ART

[0002] — Foreign language learning is experiencing strong growth all over the world, particularly in the professional sphere. Learners are looking for solutions increasingly motivating, personalized learning methods with minimal impact possible on their personal time.

[0003] There are already learning opportunities accessible on computers or mobile devices connected and using videos. Some of them stage videos to learning.

[0004] Videos provide a particularly useful audiovisual context during the communication comprehension and language acquisition. This context allows, in particular, for better understanding in which situations certain sentence structures are used (usage polite formulas etc…) and to distinguish the different accents (English American, Australian…).

[0005] — Although the video format offers various advantages in language learning, It also has some drawbacks. Generally, a video contains several minutes of dialogue, and not all the lines are interesting. adapted for the learner. It often happens that certain replies from the same some videos are too difficult, others too easy, or they are too unusual.

[0006] On the other hand, there are currently tools available to memorize more efficiently effectively using "flashcards" (micro-quizzes on micro-content) which rely on on spaced-repetition algorithms. These Algorithms facilitate the memorization process by distributing revisions of micro-content more intelligently. These algorithms optimize revision timing. of each flashcard. The more the learner masters a flashcard, the less they will need it Review often so as not to forget it.

[0007] US document 2020 / 0143704 A1 (Liao et al) published on May 7, 2020 discloses a A computer processing system is provided to enhance learning. Languages ​​via video. This system includes a video server to store videos that use one or more target languages. The system also includes a video metadata database to store translations of spoken phrases in the videos, profiles of characters appearing in the videos, and mappings between phrases and learner profiles. Furthermore, the system includes a learner profile database, a semantic analyzer, and a matching engine to find, for at least one given video and learner, alternative phrases that are sensitive to translations of the spoken phrases in that video and that conflict with a given learner profile. US document 2018 / 0293912 Al (Ni) published on October 11, 2018 discloses a method for learning English, using video clips and implementing a looping process in which video clips can be delivered over multiple time intervals over a given duration. US patent application 2021 / 0014575 Al (Selfors), published on January 14, 2021, discloses a method for learning a foreign language using video, in which subtitles can be displayed in a first target language or a second native language while a video is playing. During a pause event, both the target language subtitle and the native language subtitle can be displayed simultaneously to facilitate comprehension. While paused, a user can select an option to receive additional contextual information indicating the usage and context associated with one or more words in the target language subtitle. The user can navigate through previous and subsequent subtitles with this additional contextual information while the video is paused.This method allows users to create self-continuing video loops of definable duration, and can allow users to generate video clips by re-searching an entire database of subtitle text. Patent US2019 / 0108773 (Wyner et al.) discloses a computerized method and system for teaching a foreign language. Users can set up flashcards to study words and words in study sentences and select associated images to display with the flashcards during review. The invention relates to an improved method and system for identifying and searching for images to display. The invention also relates to improved methods for presenting flashcards for words with multiple definitions and an improved database architecture for storing flashcard data for multiple users. US patent 2014 / 0134576 Al (Edge et al.) discloses a personalized two-model language learning system and method that facilitates the learning of a new language (or a non-native language of the learner) in a personalized way. A system and process implementation defines a learner model that includes personalized information about the learner and a language model that describes language-specific information. A combination of the learner model and the language model is used to help the learner learn the language. Specifically, the learner and language models are used to create content for flashcards that are displayed to the learner. The learner's responses are used to update both the learner and language models. System and process implementations also allow the learner to play skill-based games that teach and reinforce a particular language skill the learner wishes to master. US patent 10,854,106 B2 (Geller et al.) discloses a computerized language learning system that uses targeted repetition to familiarize a student with words and rules governing the language. Targeted repetition presents practice sentences composed of specific words and rules that the language learning system has targeted for practice. The intervals between encounters with targeted words and rules vary based on previous incorrect, correct, and partially correct responses to the rule elements. Targeted reinforcement determines how to respond to the practice sentences. Learning records also track each previous response to the element, allowing the language learning system to determine which aspects of the word or rule element a student has not yet mastered.The language learning system provides targeted reinforcement by forcing the student to practice the specific aspect of the rule, as well as the practical sentence that incorporates it. This language learning system uses spaced repetition of quizzes, and the time intervals between quizzes depend on the learner's response. Patent US20120164609 (Kehoe) discloses a second language acquisition system that trains a user's auditory processing area of ​​the brain to recognize the phonemes, syllables, and words of a foreign language. This is achieved by presenting videos of a word or phrase spoken slowly and at a normal rate, and an interface that allows the user to input the phonemes, syllables, and words they have heard, with immediate feedback indicating correct or incorrect entries. This system implements a spaced repetition learning technique. The main objective of the invention is to provide a new method for more efficiently producing video content for learning a foreign language. Another objective of the invention is to provide a learning method that implements this video content. Description of the invention This objective is achieved with a process for producing micro-video content for language learning, from long-format videos extracted from one or more video content sources, implemented by computer, this process comprising the following steps: a search and identification, based on said source(s) of video content, a plurality of video content containing the same word predetermined target or the same predetermined target phrase belonging to said language, an extraction, within each of the aforementioned video content thus identified, of a video segment containing said predetermined target word or target phrase predetermined, an enrichment of each extracted video segment, by adding me- data and a contextualized test (quiz), related to the target word predetermined or said predetermined target phrase, said video segment thus enriched, constituting a micro-content video for learning. The search and identification stage can be advantageously configured to identify video content comprising a predetermined target word or predetermined target phrase corresponding to a predetermined context, and can include pre-filtering of video content according to a plurality of predetermined pre-filtering criteria, implementing an analysis of audio transcription data of said video content and / or metadata associated with said video content. Audio transcription data and / or metadata associated with video content can be delivered directly by the source of said video content. The source of video content can be advantageously provided by an external video streaming service offering an application programming interface (APD). According to another aspect of the invention, a system is proposed for producing micro-video content from long-format videos extracted from one or more video content sources, implementing the production process according to the invention, comprising: means to search for and identify, from said source(s) of video content, a plurality of video content containing the same word predetermined target or the same predetermined target phrase, means to extract, within each of said video content, thus identified, a video segment containing said predetermined target word or said predetermined phrase, methods to enrich each extracted video segment, by adding me- data and a contextualized test (quiz), related to the target word predetermined or said predetermined target phrase, so as to produce a micro-learning video content. The search and identification methods can be configured to identify video content that includes a predetermined target word or predetermined target phrase and corresponds to a predetermined context. The means for searching and identifying video content may further include means for pre-filtering video content according to a plurality of predetermined pre-filtering criteria, and means for analyzing transcription data of the audio stream of said video content and / or metadata associated with said video content. According to yet another aspect of the invention, a method for learning a language is proposed, implementing micro-video content produced by the production process according to the invention, implemented within a communication device used by a learner, this method comprising the following steps: in response to a request made by this learner based on a target word or a target phrase in that language, provision of a list of micro- video content containing this target word or target phrase and available in a remote storage space and, inputting a selection of micro-content from this list of micro- contents, viewing the selected micro-video content on the equipment of communication from the said appellant, issuing a contextualized test (quiz) contained within said micro-content video, involving the learner entering an answer, processing the input response from the audit test, implementing an algorithm of spaced repetition intended to deliver a time delay instruction for view new micro-video content associated with said target word or to said target phrase. The learning method according to the invention may further include a preliminary step to offer the learner a plurality of courses, each containing a group of micro-video content. Following yet another aspect of the invention, a language learning system is proposed, comprising: a set of micro-video content pieces, each associated with a target keyword or a target phrase in said language, each micro video content comprising a video segment previously identified as containing said target word or said target phrase, said plurality of video micro-contents being accessible from a micro-video content storage space, one or more communication devices equipped with an application a dedicated digital system implementing the learning process according to the invention, said communication equipment being connected via a communication network communication to the video micro-content storage space. The micro-content production process according to the invention thus makes it possible to generate flashcards (micro-quizzes in question / answer format) from long-form videos. A video contains many sentences, of varying difficulty. The video flashcard format allows for the isolated learning of useful sentences, at the appropriate level to facilitate learning. And above all, this format allows for more effective review thanks to a spaced repetition algorithm. These micro-quizzes are generated either manually by teachers or administrators or automatically via an algorithm by hiding, for example, the target word in the subtitle and asking the learner to recognize it in the audio track. Learning is centered on a target word, which allows the learner, for example, to discover how the same target word is used in dozens of contexts, which allows them to develop a "feeling" about that word. Furthermore, the micro-content production process according to the invention can implement an interface module with a video streaming service, enabling the indexing of large quantities of content of satisfactory quality for training purposes. This interface relies on the public API of the video streaming channel or service and uses a specific algorithm for optimal indexing within the billions of available videos. To guarantee a certain level of quality, the module includes a pre-filtering feature that rejects videos deemed unsuitable by analyzing the transcript and other metadata of the video streaming service (vulgar language, religious debates, low word count, excessively long sentences, etc.). DESCRIPTION OF THE FIGURES [Fig.1] Fig.1 schematically represents an example of an implementation of a foreign language learning system according to the invention. [Fig.2] Fig.2 represents an example of the realization of a process for producing micro-video content dedicated to learning a foreign language according to the invention. [Fig.3] Fig.3 represents an example of the implementation of a foreign language learning method according to the invention. [Fig.4] Fig.4 illustrates an example of a screenshot corresponding to a search for video micro-content by keywords. DEFINITIONS Library: a structured digital storage space for video flashcards Lesson: a set of video flashcards corresponding to a particular theme Communication equipment: smartphone, tablet, personal computer and more generally any equipment connected to one or more networks communication Video Flashcard: micro-content compatible with online learning smartphone, tablet or personal computer, through spaced repetitions Indexer: a digital tool for searching for videos containing a word or phrase. target pressure Quiz: a test or questionnaire used to test specific knowledge or skills Spaced repetition: a memorization technique based on neuroscience which consists of showing the element (for example, a quiz) again to memorize just before the learner forgets DETAILED DESCRIPTION A foreign language learning system 1 according to the invention comprises, with reference to [Fig.1], a set of personal communication equipment C1,C2,,...Ck used by a plurality of learners S1,S2,...Sk and connected via a communication network 30 to a remote Cloud-type storage space 33. Each personal communication device C1,C2,...CK received a software application comprising a graphical user interface ("front end") dedicated to the learning process according to the invention, by downloading from one or more application operators. A system 2 for producing video flashcard micro-content according to the invention includes a dedicated server 34 (“back end”), provided with a storage space 32 and connected, via the communication network 30, to a video chain 11 and to the Cloud 33 which contains a database dedicated to storing the micro-content thus produced. Production system 2 is dedicated to the generation of video flashcards 7.1,7.2;..7.a each associated with a target word or expression detected in a video segment 4.1,4.2,..4.n extracted from video content broadcast by video channel 11. These video flashcards 7.1, 7.2, ..., 7.n can be categorized into courses 9.1, 9.2, ..., 9.n, which are made available to learners. In practice, these video flashcards and courses are stored in a library L within the cloud 33. Each learner S1, S2, ..., Sk can consult a course catalog on the learning server 30. available and order one corresponding to their learning objective. This learner will then receive on their communication device a set of flashcards corresponding to this course, in a sequential manner, which will be returned for review via spaced repetition. It should be noted that the use of courses is not mandatory and that both learners and teachers can also search for video contexts based on the requested keywords. The use of courses is therefore an optional feature that adds pedagogical value by allowing teachers to better organize flashcards to make learning more meaningful. We will now describe in more detail, with reference to [Fig. 2], an example of the implementation of the flashcard production method according to the invention. This production method 10 comprises a continuous search 12 for videos containing target words or phrases on a video streaming channel. It should be noted that this search 12 can implement an API of the video streaming channel. Videos detected on the basis of target words or target phrases are indexed (3) to produce video segments 4 associated with the target words or phrases, which are subjected to a pre-filtering algorithm 6, for example of the NLP type for “Natural Language Processing”. From the indexed and pre-filtered video segments, video flashcards (also known as video cards or video micro-content) are created (13), each containing a quiz in the form of a set of questions related to the content of the video segments and their connection to the target word or phrase. This quiz within the video flashcard can be in audio and / or script format, based on the video content of the flashcard. These quizzes are generated either manually by teachers or administrators or automatically, for example, by asking the learner to complete a translation exercise. The video flashcards thus created can be stored in the dedicated library L within the cloud 33. They can also be pre-organized (8) in courses 9.1,9.2,9.3 — for example “communicating by telephone”, “food vocabulary” or “vocabulary of places” — within the library L stored in the cloud 33. Each of these courses 9.1,9.2,9.3 contains a plurality of video flashcards 7.1,7.2,7.3 dedicated to the theme of the corresponding course. We will now describe, with reference to [Fig. 3], an example of the implementation of the learning process according to the invention. This learning process 20 includes an initial step 21 to allow a user learner S – or a teacher – to select a flashcard either directly (as illustrated by [Fig. 4] described below) from the library L based on a query for a target word or target expression, either by selecting a course from an offer of courses 9.1,9.2,9.3 available in the L video flashcard library. With reference to [Fig.4], the software application downloaded into the communication equipment of a learner S provides him with a graphical user interface 40 allowing him to directly search for a video flashcard of interest in the library L stored in the cloud 33, by entering a target word (for example "PATENT") or a target phrase in an input window 41 and acting on a SEARCH button or icon 42. In response to the query "PATENT", a list of relevant video flashcards 43, 44, 45 is displayed, along with excerpts of indexing content corresponding to each listed video segment. Learner S can then select a flashcard from this list to view. The selection step 21 is followed by a step 22 of viewing a video flashcard thus selected on the communication equipment used by the learner S, and then by a step 26 of answering a quiz or test contained in this video flashcard. The learner S is thus invited (26) to answer one or more questions related to the target content of the video flashcard. The answer entered by the learner S—which can be in several formats: voice, alphanumeric, or selection of an answer from a set of possible answers—is processed (24) by the "front end" software application previously downloaded onto the communication equipment used by the caller. A spaced repetition algorithm is implemented within the software application dedicated to the process. For the practical implementation of this algorithm, it is useful to refer to the state of the art in the field of spaced repetition system (SRS) learning. In particular, the document "Adaptive Spacing Algorithms for Learning to Optimize Long-Term Mastery of Knowledge Components" by Benoît Choffin, Université Paris-Saclay, 2021. NNT: 2021-UPASG001-03216648, can be cited. This spaced repetition algorithm 23 automatically determines the time intervals between micro-content releases and between quiz response requests, based on previous responses. The caller S is invited to retake the quiz after a time interval ÂT which grows indefinitely (25). If the caller S fails in this quiz, he is invited (27) to retake the quiz one minute later (the interval is reset). As a non-limiting example, it can be predicted that if the caller has answered correctly twenty times, the AT time interval will exceed one year. In practice, users stop using the system before then. With each review, the learner's memorization is reinforced. In practice, memorization is considered almost stable when the interval is greater than ten days. Of course, the invention is not limited to the embodiments just described, and many other configurations can be considered without departing from the scope of the present invention. In particular, other configurations for processing quiz responses than the one just described can be implemented.

Claims

Demands

1. Method (10) for producing video micro-content (7; 7.1, 7.2, 7.3) intended for language learning, using videos in the format long extracts from one or more video content sources (11), put in computer-generated work, this process comprising the following steps: a search and identification (12), from the said or said sources of video content (11), of a plurality of video content containing the same predetermined target word or a predetermined target phrase belonging to said language an extraction, within each of said video content thus identified, from a video segment (4) containing said target word predetermined or said predetermined target phrase, an enrichment of each video segment (4) thus extracted, by adding metadata and a contextualized test (quiz), in relationship with said predetermined target word or phrase predetermined target, said video segment thus enriched constituting a micro video learning content (7).

2. Production method {10) according to the preceding claim, ca- characterized in that the search step is configured to identify video content including a predetermined target word or a predetermined target phrase corresponding to a predetermined context. Claim OC: Production according to any one of the preceding claims. Claim 3: Production method according to any one of the preceding claims. previous, characterized in that the video content identification step includes pre-filtering of video content according to a plurality of predetermined pre-filtering criteria, implementing an analysis of audio transcription data of said video content and / or me- data associated with said video content.

4. A production method according to the preceding claim, characterized in what the audio transcription data and / or metadata associated with video content are delivered directly by the source of said video content.

5. A production method according to any one of the preceding claims. previous, characterized in that the source of video content is procured through an external video streaming service offering an interface of application programming (API.

6. System (2) for producing video micro-content (7.1, 7.2, ..., 7n) at starting from long-format videos extracted from one or more sources of video content (11), implementing the production process according to any one of the preceding claims, including: means to identify, from the said or the said sources of video content, a plurality of video content containing the same predetermined target word or the same predetermined target phrase, means to extract, within each of said contents videos thus identified, a video segment containing said word predetermined target or predetermined phrase, methods to enrich each extracted video segment, by adding metadata and a contextualized test, in relation with the said predetermined target word or predetermined target phrase- finished, so as to produce a micro video content learning.

7. Production system (2) according to the preceding claim, characterized in that the means of identification are configured to identify video content including a predetermined target word or phrase predetermined target corresponding to a predetermined context.

8. Production system (2) according to any one of the two claims previous indications, characterized in that the means of research and video content identification methods also include means to pre-filter video content according to a plurality of pre-defined criteria completed pre-filtering, and means for analyzing data from transcription of the audio stream of said video content and / or me- data associated with said video content.

9. Method (20) for learning a language, implementing micro- video content (7.1, 7.2, …7n) produced by the production process according to any one of claims 1 to 5, implemented within of a communication device used by a learner, this process including the following steps: in response to a request made by this learner on the database of a target word or target phrase in said language, provision of a list of micro-video content containing this target word or target phrase and available in a space remote storage and, entering a selection of micro-content from this list of micro-content, viewing the selected micro-video content, on the communication equipment of the said appellant, issuance of a contextualized test contained in said micro- video content, involving the input of a response by the learner, processing of the response entered to said request, putting in implements a spaced repetition algorithm designed to deliver a time delay instruction to view a new micro-video content associated with said target word or phrase target.

10. A learning method according to the preceding claim, characterized in that it also includes a preliminary step to propose to the learner a plurality of courses each containing a group of micro- video content.

11. | A language learning system, comprising: a set of micro-video content pieces, each associated with a target word or target phrase in said language, each micro-video content including a pre-video segment clearly identified as containing said target word or said target phrase, the said plurality of micro-video content being ac- transferable from a micro-content storage space video, one or more communication devices equipped with dedicated digital application implementing the process learning according to one of the two pre- claims previous, the said communication equipment being connected via a communication network to the space of storage of micro video content.