Classifications

• • G—PHYSICS
  • G09—EDUCATION; CRYPTOGRAPHY; DISPLAY; ADVERTISING; SEALS
  • G09B—EDUCATIONAL OR DEMONSTRATION APPLIANCES; APPLIANCES FOR TEACHING, OR COMMUNICATING WITH, THE BLIND, DEAF OR MUTE; MODELS; PLANETARIA; GLOBES; MAPS; DIAGRAMS
  • G09B19/00—Teaching not covered by other main groups of this subclass
  • G09B19/0053—Computers, e.g. programming
• • G—PHYSICS
  • G09—EDUCATION; CRYPTOGRAPHY; DISPLAY; ADVERTISING; SEALS
  • G09B—EDUCATIONAL OR DEMONSTRATION APPLIANCES; APPLIANCES FOR TEACHING, OR COMMUNICATING WITH, THE BLIND, DEAF OR MUTE; MODELS; PLANETARIA; GLOBES; MAPS; DIAGRAMS
  • G09B7/00—Electrically-operated teaching apparatus or devices working with questions and answers

Definitions

• TITLE A method of assisting in the learning of a plurality of information items by a user of a terminal.
• the field of the invention is that of learning assistance.
• the invention relates to a method of assisting in the learning of information by a user of a terminal.
• information is meant in particular, but not exclusively, the names of technologies, concepts, places, people, etc., or any other information (also called “knowledge” or “competence”) which may be useful to the organization. 'user.
• terminal is meant in particular, but not exclusively, a personal computer (fixed or portable), a digital tablet, a personal digital assistant, a smartphone, a workstation, etc., or any device other than a user. can use to receive, send or search text and / or image and / or sound type content.
• content is meant in particular, but not exclusively, an electronic mail, a message (instantaneous or not), a document, a search (carried out for example with a web browser), a news feed of a social network , content published on a social network, etc.
• the invention can be applied in many fields, for example:
• the proposed solution allows the user of a social network to be informed, by real time or in deferred time, as soon as new information (new concept or new element), never encountered previously, appears in the news feed of his social network);
• the invention in at least one embodiment, aims in particular to overcome these various drawbacks of the state of the art.
• an objective is to provide a technical solution to aid in the learning of information by the user of a terminal.
• At least one embodiment of the invention also aims to provide such a solution which is simple to implement and easy to use.
• Another objective of at least one embodiment of the invention is to provide such a solution which makes it possible to limit the computing resources of the computing machine, as well as the network traffic to and / or from the terminal of the. user.
• a complementary objective of at least one embodiment of the invention is to provide such a solution which makes it possible to adapt both to the user and to the information to be learned.
• a method for assisting in the learning of a plurality of information items by a user of a terminal.
• the method comprises, during a user encounter with a given item of information, from among the plurality of items of information, during use of the terminal: storing in a knowledge base, context data relating to said encounter, with an entry for the given information; o determination of a knowledge index of the given information, specific to the user, as a function of context data recorded in the knowledge base with the entry for the given information; and o proposal for access to at least one element for understanding the information given as a function of the determined knowledge index.
• the proposed solution offers a completely new and inventive approach, consisting of a learning aid method which is implemented in a computing machine. It aims to identify the information which may be unknown, or quite simply forgotten, by the employee (but which may be important, for example in the nature of his profession) by determining an index (or degree) of knowledge for each of this information.
• the proposed solution therefore aims to provide the user with elements of understanding based on the knowledge index determined. Thus, if the knowledge index is deemed too low (for example because it is below a threshold), elements of understanding are offered to the user while if the knowledge index is high, no element of understanding is not offered.
• an advantage of the proposed solution is that it makes it possible to limit the computing resources used by the computing machine, as well as the network traffic to and / or from the user's terminal, since the number of notifications, for offer the user access to the elements of comprehension then to provide these elements of comprehension (if the user wishes it), is limited (the calculation machine automatically selects the information for which it is necessary to propose elements understanding).
• the element of comprehension proposed is a complete initial training, while that for a value of the index close to a threshold (for example 0.45 "if the threshold is equal to 0.5) and indicating that the information given is possibly forgotten by the user, the comprehension element proposed is for example a shorter training (or a simple reminder of the definition).
• One advantage of the proposed solution is that it is easy to implement since it suffices, in addition to the terminal already available to the user, a computing machine (possibly the one already present in the terminal) cooperating with a knowledge base (database).
• Another advantage of the proposed solution is that it is easy to use since the user is offered access to elements of understanding, for information that the calculation machine itself has automatically selected, according to of the knowledge index determined.
• Another advantage of the proposed solution is that it makes it possible to adapt both to the user (since the knowledge index is a function of the content of the knowledge base, which itself depends on the choices of the user. user to access or not to the elements of understanding offered) and to the information to be learned (since the knowledge index is specific to a given information; in other words, each of the information is associated with its own knowledge index ).
• the proposed solution makes it possible to take into account the facilities or, on the contrary, the difficulties that each user may encounter in particular skill areas (that is to say for particular information to be learned).
• the method comprises storing, with said entry, at least one piece of context data relating to said access. , respectively audit non-access.
• the encounter of the user with the given information belongs to the group comprising:
• the proposed solution can take account of the great diversity in the encounters that the user can have with a given item of information. It is effective even if the user manipulates a large amount of information.
• the list of types of meetings is not exhaustive.
• said at least one comprehension element belongs to the group comprising:
• said at least one piece of information data belongs to the group comprising:
• the context data belong to the group comprising:
• the determination of the knowledge index depends on:
• the knowledge index can be calculated easily during a certain number of first iterations of the learning aid process and as long as the knowledge base is not sufficiently filled for a calculation of d 'index based on a machine learning model (see detail below) is considered acceptable.
• the determination of the knowledge index uses a machine learning model and comprises:
• the knowledge index can be calculated efficiently and based on several criteria (corresponding to the different information data and context).
• the attributes included in the input data belong to the group comprising:
• At least one context attribute filled in with the context data and belonging to the group comprising: a reference date, defined as the most recent date among one or more date (s) of encounter of the given information and a or more date (s) of proposal for access to at least one element of understanding; a number of encounters of the information given in reading in a predetermined period preceding the reference date; a number of encounters of the information given in writing in said predetermined period; a number of searches for the information given, by the user, in said predetermined period; an average number of sentences in contents in which the user has encountered the information given in reading in said predetermined period; an average number of sentences in contents in which G user has encountered the information given in writing in said predetermined period; and a number of accesses to said at least one comprehension element, in said predetermined period.
• the method comprises a construction of the machine learning model, by performing a determined number of construction iterations, each corresponding to an iteration of the learning assistance method, and each comprising the following steps:
• the machine learning model can be built using the information and context data stored during certain iterations of the process, taking into account in particular whether or not the user has access to the elements of understanding.
• the estimate of the knowledge index is equal to:
• the first value is “0” and the second is “1".
• the construction of the machine learning model is carried out again after a predetermined number of iterations of the learning assistance method and / or at a predetermined frequency.
• the machine learning model can evolve over time, to improve learning as the number of iterations of the process grows, that is, as the knowledge base content increases.
• a computer program product comprising program code instructions which, when they are executed by a computing machine, cause the computing machine to perform the aforementioned method. (in any of its various embodiments).
• a non-transient, computer readable storage medium storing the aforementioned computer program product.
• a computing machine configured to perform the above method (in any of its various embodiments).
• FIG. 1 shows a simplified flowchart of the method according to the invention
• FIG. 2 is an example of a forgetting curve, used in some iterations of step E5 in Figure 1;
• FIG. 3 shows a simplified flowchart of the construction of the machine learning model used in some iterations of step E5 of Figure 1;
• FIG. 4 shows the structure of a computing machine, according to a particular embodiment, configured to carry out the method of FIG. 1.
• the method is implemented by a computing machine (also called a “system” in the remainder of the description), an exemplary structure of which is presented below, in relation to FIG. 4.
• the computing machine implementing the method is integrated into, or confused with, the user's terminal (which is for example a fixed or portable personal computer, a digital tablet, a personal digital assistant, a smartphone, a workstation, etc. ).
• the computing machine implementing the method is integrated into, or merged with, another device which cooperates with the user's terminal (such as a home gateway, also called an "Internet box").
• a step E1 the computing machine scans, thanks to one or more probes, the user's activity on the terminal, comprising for example the contents of text and / or image and / or sound type that the user has received, sent or searched.
• content is understood in particular, but not exclusively, an electronic mail, a message (instantaneous or not), a document, a search (carried out for example with a web browser), a thread of news from a social network, content published on a social network, etc.
• GDPR general data protection regulations
• a step E2 the computing machine analyzes the scanned content and attempts to extract therefrom information to be learned by the user.
• the extraction is based for example on referential linguistic expressions (named entities) and simple and extended phrases.
• this information also called “key elements” are for example the names of technologies, concepts, places, people, etc., or any other information (also called “knowledge” or “skill”) that may be of use to the user.
• the computing machine detects an encounter of the user with one or more information (key elements), during use of the terminal.
• information key elements
• user encounter with given information we mean, for example, the presence of the information given in:
• a test step E3 for a given item of information extracted in step E2 (that is to say for a meeting of the user with this given item of information), the computing machine determines whether there is already an entry for the information given in a knowledge base 1 (database) aggregating information (key elements) for the user. For this, the computing machine queries the knowledge base 1, as symbolized by the arrow referenced 2.
• step E3 If no entry for the given information exists in the knowledge base (negative response to the test of step E3), the algorithm goes to step E7 in which the computing machine creates in the knowledge base 1 an entry for the given information, and stores with this entry at least one piece of information data relating to the given information and at least one context data item relating to the meeting.
• step E7 the computing machine stores for example:
• Step E7 is followed by step E8 in which the computing machine offers the user access to (at least) one element of understanding the information given.
• the element of comprehension is chosen based on a value of the knowledge index.
• element of comprehension is meant for example: a definition of the given information, an explanation of the given information, training on the given information, help relating to the given information, a learning element (written and / or oral and / or visual) of the information given, etc.
• the comprehension element is structured in such a way that it comprises basic information which is supplemented or enriched according to a defined tree structure, the elements of this tree structure being selectable according to the value of the index of awareness.
• the comprehension element can be structured by information of different size or quantity of data, in a data matrix for example.
• This matrix is constructed by taking into account, for example, characteristics of the duration of reading of this element of comprehension or of complexity.
• a comprehension element of greater or lesser reading time or of greater or less complexity may be selected from this matrix.
• step E3 If an entry for the given information already exists in the knowledge base (positive response to the test of step E3), the algorithm goes to step E4 in which the computing machine stores in the knowledge base, with the existing entry, at least one (other) piece of context data relating to the meeting.
• the computing machine calculates an index of knowledge (by the user) of the information given, according to the content of the knowledge base.
• one and / or the other of two methods is used, for example, depending on the number of iterations of the process already carried out before the current iteration.
• N 1000
• the first method includes, for example, a calculation of the knowledge index based on:
• a forgetting curve such as for example the Ebbinghaus curve 21 illustrated in FIG. 2. with the time on the abscissa and the retention percentage on the ordinate; the curve referenced 22 corresponds to the case where the user is reminded of the information at the various times mentioned on the abscissa (10 min, 1 day, 1 week, 1 month and 6 months); the double arrow referenced 23 illustrates the gain obtained after six months (that is to say the difference between the two aforementioned curves 21 and 22).
• the second method includes, for example, the following steps to calculate the knowledge index:
• the input data includes the following attributes:
• context attributes calculated (filled in) with the context data such as for example the following attributes: o reference date, defined as the most recent date among the dates stored in the knowledge base for the given information (date ( s) of the user's encounter with the given information and date (s) of access or non-access to the element of understanding); o number of encounters of the information given in reading in a predetermined period preceding the reference date; o number of encounters with the information given in writing during the aforementioned period; o number of searches for the information given, by the user, in the aforementioned period; o average number of sentences in contents in which the user has encountered the information given in reading in the aforementioned period; o average number of sentences in contents in which the user has encountered the information given in writing during the aforementioned period; o number of accesses to the comprehension element, in the aforementioned period; o etc.
• Step E5 is followed by a test step E6, in which the computing machine compares the knowledge index with a predetermined threshold. If the knowledge index is greater than or equal to the threshold, the algorithm returns to step El, for a new iteration of the process. If the knowledge index is below the threshold, the algorithm goes to step E8 already explained above (proposal for access to at least one element of understanding).
• Step E8 is followed by a test step E9, in which the computing machine determines whether the user has accessed the understanding element.
• step E10 the computing machine stores in the knowledge base (as symbolized by the arrow referenced 3), with the existing entry, at least one other context data item , relating to access.
• step E10 the algorithm returns to step El, for a new iteration of the method.
• step El i the computing machine stores in the knowledge base (as symbolized by the arrow referenced 4), with the existing entry, at least one other piece of data of context, relating to non-access.
• step El i the algorithm returns to step El, for a new iteration of the process.
• FIG 3 shows a simplified flowchart of the construction of the machine learning model, which model is used in some iterations of step E5 of Figure 1, as discussed above.
• the construction of the model comprises a determined number M of construction iterations, each corresponding to one of the iterations of the learning aid method of Figure 1.
• the computing machine In a step 31, the computing machine generates an input data item as defined above (see step E5 of FIG. 1), that is to say comprising the plurality of attributes (themselves determined as a function of information and context data stored, with the entry for the given information, in the knowledge base).
• the computing machine determines an estimate of the knowledge index, as a function of the user's access or non-access to the comprehension element (see test step E9 of FIG. 1 ).
• the estimate is equal to a first value (for example "0") indicating that the given information is not known to the user, in the event of the user having access to the comprehension element. (positive response to the test step E9), and to a second value (for example "1") indicating that the given information is known to the user, in the event of non-access by the user to the element of understanding (negative response to test step E9).
• the computing machine provides the machine learning model with the input data, along with the known result (defined as the estimate of the knowledge index, calculated in step 32).
• a test step 34 the computing machine determines whether the number M of construction iterations has been performed. If not, the algorithm returns to step 31, for a new iteration of construction. If so, the algorithm proceeds to finish step 35.
• the construction method of Figure 3 can be performed again after a predetermined number of iterations of the method of Figure 1 (for example after the iterations M + 1 to 2M of Figure 1, then after the iterations 2M + 1 to 3M in Figure 1, and so on).
• the construction process of Figure 3 can also be performed again at a predetermined frequency (eg once a week).
• FIG. 4 presents an example of the structure of a computing machine 40 for carrying out (executing) the method of FIG. 1.
• This structure comprises a random access memory 42 (for example a RAM memory), a read only memory 43 (for example a ROM memory or a hard disk) and a processing unit 41 (equipped for example with a processor, and controlled by a program. computer 430 stored in read only memory 43).
• the code instructions of the computer program 430 are for example loaded into the random access memory 42 before being executed by the processor of the processing unit 41.
• FIG. 4 illustrates only one particular way, among several possible, of implementing a computing machine to carry out (execute) the method.
• the computing machine can be implemented indifferently in the form of a reprogrammable computing machine (a PC computer, a DSP processor or a microcontroller) executing a program comprising a sequence of instructions, or in the form of a dedicated computing machine (for example a set of logic gates such as an FPGA or an ASIC, or any other hardware module).
• a reprogrammable computing machine a PC computer, a DSP processor or a microcontroller
• a program comprising a sequence of instructions
• a dedicated computing machine for example a set of logic gates such as an FPGA or an ASIC, or any other hardware module.
• the corresponding program (that is to say the sequence of instructions) could be stored in a removable storage medium (such as for example a diskette, CD-ROM or DVD-ROM) or not, this storage medium being partially or totally readable by a computer or a processor.
• a removable storage medium such as for example a diskette, CD-ROM or DVD-ROM
• an attempt is made to resolve the following problem: how to identify / detect information not known or not well known (which may be forgotten) of the employee with regard to several criteria (for example, the time between a determined number of appearances (encounters) of this same information in its activities, the nature or the field associated with said information, ).
• an employee of a company may receive an electronic message (e-mail) of such a nature as to contain: "For that you have to do a DMI".
• DMI meaning “Request for Computing Means”
• the learning aid process (see figure 1) will detect that the acronym DMI is not information known to the employee in question, and will be able to provide him with one or more elements of understanding (in this case an access to the tool allowing to formulate a request for IT resources).
• the process aggregates in a database information that is the subject of learning (key elements) such as: names of technologies, concepts, locations, people or any other information that the employee encounters in his daily activities (e-mails , conversations on messaging (instantaneous or not), documents ... written or read) for a more or less long period and configurable according to his preferences or his profession.
• key elements such as: names of technologies, concepts, locations, people or any other information that the employee encounters in his daily activities (e-mails , conversations on messaging (instantaneous or not), documents ... written or read) for a more or less long period and configurable according to his preferences or his profession.
• the system By analyzing the employee's inputs over time such as the elements (content) he receives: e-mails, conversations, new documents read, the system detects the information contained in these entries and checks for the presence of this information. in the employee's knowledge base. If information detected in one of the content is not present in the knowledge base: the system considers this information to be potentially new, and therefore not known to the employee.
• the system will calculate a knowledge index. This index is based on several criteria, for example: the time between two appearances of the information (which may suggest forgetting if this time is long), the domain associated with this information (network, project management, new technologies, etc. ), the nature of the information (IS tools, contact, organization, etc.), the context in which the information appears, etc.
• the index will be calculated on the observation of the knowledge base of the employee and, in a particular embodiment, on the observation by the system of requests by the employee of additional information. (or on the contrary the absence of request for additional information) for domain information or of a similar nature.
• the knowledge index is therefore specific to the information and specific to the user.
• the knowledge base can be initialized by information to which the employee has given access over a given period. Then the index will be based on average memorization assumptions, such as those that can be seen on the Ebbinghaus Curve.
• step El Consider the information (acronym) "APN" that the employee reads in an e-mail (step El). It is not present (step E3) in the knowledge base which contains information corresponding for example to one month (initialization from information from the history of e-mails, instant messages or others).
• the system then stores (step E7) the following data (non-exhaustive list) in the knowledge base:
• the system displays (E8) a proposal to access the definition of this acronym (information) to the user.
• the system extrapolates and assumes that the user knows this information. It adds (Eli) to the knowledge base the following new context data (in particular context data indicating "Known information"), which will make it possible to produce the specific learning model associated with the employee:
• the knowledge base contains too little information and data to extrapolate and calculate the knowledge index of the employee's system. He will therefore calculate (E5) the knowledge index on the basis of the Ebbinghaus curve.
• step E3 If the user searches for the definition of information himself (acronym "APN" (El), the system will detect it (E2). If the information is not present (step E3) in the knowledge base, the system then stores (step E7) the following data in the knowledge base:
• the system displays (E8) a proposal to access the definition of this acronym (information) to the user. See paragraph A above for the rest (depending on whether the user clicks on the definition or not, going to step E10 or Eli).
• the knowledge base will thus be filled with information and data specific to the user and allow the production of a model capable of assigning a knowledge index to each item of information.
• the data in the knowledge base is analyzed and allows the production of input data, for a classic process of construction (training) of the machine learning model.
• Each input data item is accompanied by the associated known result (knowledge index is equal to “1” if the information is known or to "0" if the information is not known.
• the construction algorithm then identifies trends in this input data and results, which will match the attributes of the input data to the result (target, i.e. the value of the knowledge index to predict, and it outputs a machine learning model that captures those trends.
• the attributes (characteristics) to be taken into account to produce input data for the machine learning model are for example (non-exhaustive list):
• the knowledge index is in the process the result, that is to say the qualitative variable to be predicted by learning.
• Each input data item produced on analysis of the knowledge base is for example vectorized.
• Each value of one of the attributes (characteristics) of an input data is reduced to a digital data to produce a vectorized input data.
• AI classification-type artificial intelligence
• the model will make it possible to obtain the knowledge index, a real value varying from 0 (which means that the information is not known) to 1 (which means that the information is known) for a new vectorized input data item making it possible to answer the question (with values entered for indications A to L): "what is the knowledge index of an item of information with label A, of domain B, of nature C, of type D, for which the last appearance (encounter) took place between E and F months, for which the number of appearances in reading took place G times, in writing H times, in a context of the appearance of more than I sentences in reading and J sentences in writing, for which definitions were read K times in the month preceding the last occurrence ...? ".
• the chosen construction algorithm depends for example (“deep leaming”, “LSTM

Landscapes

• Engineering & Computer Science (AREA)
• Theoretical Computer Science (AREA)
• Business, Economics & Management (AREA)
• Physics & Mathematics (AREA)
• Educational Administration (AREA)
• Educational Technology (AREA)
• General Physics & Mathematics (AREA)
• Computer Hardware Design (AREA)
• Entrepreneurship & Innovation (AREA)
• Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Applications Claiming Priority (2)

Publications (1)

Family

ID=73497846

Family Applications (1)

Country Status (4)

Families Citing this family (2)

Citations (2)

Family Cites Families (11)

• 2020
  • 2020-06-25 FR FR2006657A patent/FR3111728A1/fr not_active Withdrawn
• 2021
  • 2021-06-17 WO PCT/FR2021/051095 patent/WO2021260299A1/fr not_active Ceased
  • 2021-06-17 EP EP21740596.8A patent/EP4172973A1/de active Pending
  • 2021-06-17 US US18/003,305 patent/US20230245590A1/en active Pending

Patent Citations (2)

Non-Patent Citations (1)

Also Published As

Similar Documents

Legal Events