PRIORITY STATEMENT

[001] The present application hereby claims priority from Indian patent application with the application number 202241042572, filed on 25 July 2022, the entire contents of which are incorporated herein by reference.

FIELD OF TECHNOLOGY

[002] The present disclosure generally relates to data analytics and machine learning and more particularly to a machine learning (ML) based learning system for estimating a deficit in the knowledge required for completing a task presented to at least one user and a method thereof.

BACKGROUND

[003] Today’s classroom-based education system is basically passive or didactic, where a teacher narrates or displays the teaching material on board, and the students listen. In such type of classroom setting, it is observed that there is no synchronization of the student with the teacher. The teacher also does not have the time to check on the progress of each student in the classroom. This lack of classroom teaching effectiveness has resulted in the mushrooming of tuition centres and private tutors, where most of the students go to compensate for their deficient studies, despite spending a long time every day in classrooms. In spite of the teachers being qualified, the school managements being sincere, and the students being smart, it is the system which is failing them.

[004] Due to the existing learning system infrastructure, there is a possibility of having knowledge gaps and its widening with every academic year, since the classroom education model is built in such a way that every student is moving forward at the same pace as in an assembly line. The students are grouped by age and assigned to courses with similar lengths of time and rigidly prescribed curricula. They all learn the subject material at the same rate, and lectures are used to saturate students with information. A majority of people have thought, "Well, I should have paid more attention in previous classes", upon entering a senior mathematics class. These fundamental inadequacies cannot possibly be filled in a classroom setting.

[005] Meeting every student's individual learning need is a crucial component of the educational system. Depending on their personality traits and abilities, people learn in different ways. For instance, while teaching a subject in a classroom setting, teachers frequently employ a range of instructional tactics in the hopes that one of them would correspond to each student's preferred learning style. However, due to time restrictions, it might be challenging for a teacher to devote all of their attention to a student who is struggling to learn a concept or skill due to time constraints. As a result, the teacher conducts a test that is marked and recorded in order to monitor student development and curricular advancement. The teacher then moves on to teaching the following lesson in the curriculum following the test. Test results, however, might not be able to identify any areas where a student's knowledge is lacking or track their development.

[006] Now, there are innumerable online learning platforms for the students to learn at their own place and pace. The online learning platforms have solved the problem of understanding the subject matter effectively by hosting the relevant content such as training videos, documents, presentations, and lecture notes on-line. The state-of-the-art online learning platforms mostly have academic content hosted. The online learning platforms have limited effectiveness as there is almost no support for problem solving. Many online platforms now have online tutors to address this problem. But the issue with online tutors is that they cannot be available throughout the day, any day of the week. This lack of immediate support in problem solving is frustrating to the users.

[007] The United States patent titled, “adaptive learning machine for score improvement and parts thereof’ bearing the patent number US10,854,099 B2, assigned to Indiavidual Learning Ltd., provides a system and method to provide a self-leaming/adapting systems by generating challenges based on user’s previous attempts and other factors, generating remedial recommendations, such as, “you spend too much time on questions where you are unsure of the subject matter”, or “you spend too little time reading all the questions before you start attempting”.

[008] There are several other disclosures that teach identifying knowledge gaps and bridging them but somehow or the other, they are either not advanced enough or not ideal for holistic learning purposes. United States Patent numbered US11416558B2 titled, “System and method for recommending personalized content using contextualized knowledge base”, also assigned to Indiavidual Learning Ltd., for instance, relies on standard node-based knowledge database where each knowledge is represented as nodes and thus traced. It is a widely commercially available solution but is majorly used for setting up question papers. Similarly, United States published application numbered US20120040326 Al and titled “Methods and systems for optimizing individualized instruction and assessment” and the journal article titled “Effects of an integrated concept mapping and web-based problem-solving approach on students' learning achievements, perceptions and cognitive loads” by Gwo-Jen Hwang, et al, (Computers & Education, Volume 71, February 2014, Pages 77-86) look at the problem with statistical approaches for programming a set standard database based on user records where knowledge is considered a super set. United States patent numbered US9626875B2, titled, “System, device, and method of adaptive teaching and learning”, assigned to Time to Know Ltd., is another approach where the invention does not take into consideration pattern combination into account but sets a “required knowledge map” that students are expected to meet, by grouping and regrouping their analytics, thus ignoring past knowledge gaps.

[009] In yet another published United States patent application numbered US20060154226A1, titled “Learning support systems” assigned to Perpetual Improvement Lie., discloses a behaviour optimized problem solving where the focus is more on psychological aspects, especially as it addresses the time taken by a learner to solve a problem rather than being driven by data. The paper titled “Concept-Aware Deep Knowledge Tracing and Exercise Recommendation in an Online Learning System” by Fangzhe Ai et al, in Proceedings of the 12 ^th International Conference on Educational Data Mining (EDM 2019), pages 240 to 245, explores the use of Artificial Intelligence and Machine Learning in this field. It, however, remains inadequate as the system is non-linear owing to instantaneous analysis. Although it attempts deep reinforcement learning (DRL) it does not support the building blocks of learning. Consequently, it does not make learning holistically sound, and the problem persists.

SUMMARY

[0010] This summary is provided to introduce concepts of the subject matter in a simple manner that is further described in the detailed description of the disclosure. This summary is not intended to identify key or essential inventive concepts of the subject matter nor is it intended to determine the scope of the disclosure.

[0011] To overcome at least one of the problems in the state of the art, that is, presence of cascading and cumulative knowledge gaps in students that causes difficulties in completing any task, a system and method for estimating a deficit in knowledge required for completing a task presented to at least one user (also termed as knowledge gap) and filling the knowledge gap of at least one user for completing a task is needed. The problems with these cascading and cumulative knowledge gaps and corresponding problem-solving gaps, in a classroom teaching is that most of the students in a class have practically no chance to master the subject matter, once they slip-up in their previous classes (years). To make matters worse, for many students, it is psychologically damaging when they are classified as average or poor students, whereas the fault lies in the underlying didactic classroom teaching system.

[0012] Briefly, according to an exemplary embodiment, a system for estimating a deficit in knowledge required for completing a task presented to at least one user is disclosed. The system includes a user device configured for collecting a first data based on an interaction of the at least one user with the user device while completing the task presented to the at least one user on the user device. The system includes a data communication device of the user device configured for transmitting the first data to a server for analysing the first data based on a first set of parameters and a second set of parameters. The system includes the server configured for receiving the first data and a second data. The server includes a model which is trained iteratively, using the second data. The second data being the data based on each interaction of each user of a plurality of users while completing each task of a plurality of tasks, associated with an academic subject, presented to each of the plurality of users. The model is trained for determining a familiarity score for the at least one user indicative of familiarity of the user with the academic subject based on a learning history of the user. The model is trained for determining a competence score for the at least one user, based on analysis of the first data; and estimating the deficit in knowledge required for completing a task presented to the at least one user, using the output of the model and the first data.

[0013] Briefly, according to an exemplary embodiment, a method for estimating a deficit in knowledge required for completing a task presented to at least one user is disclosed. The method includes collecting a first data using a user device, based on an interaction of the at least one user with the user device while completing the task presented to the at least one user on the user device. The method includes transmitting the first data to a server using a data communication device of the user device for analysing the first data based on a first set of parameters and a second set of parameters. The method includes receiving the first data and a second data by the server for training a model iteratively, using a second data, the second data being the data based on each interaction of each user of a plurality of users while completing each task of a plurality of tasks, associated with an academic subject, presented to each of the plurality of users. The model is trained for determining a familiarity score for the at least one user indicative of familiarity of the user with the academic subject based on a learning history of the user; determining a competence score for the at least one user, based on analysis of the first data; and estimating the deficit in knowledge required for completing a task presented to the at least one user, using the output of the model and the first data.

[0014] The summary above is illustrative only and is not intended to be limiting in any way. In addition to the illustrative aspects, exemplary embodiments, and features described above, further aspects, exemplary embodiments, and features will become apparent by reference to the drawings and the following detailed description.

BRIEF DESCRIPTION OF THE FIGURES

[0015] These and other features, aspects, and advantages of the exemplary embodiments can be better understood when the following detailed description is read with reference to the accompanying drawings in which like characters represent like parts throughout the drawings, wherein:

[0016] Figure 1 illustrates a high-level block diagram of a machine learning (ML) based system 100 configured for ‘building knowledge’ for concepts-based problem solving in a specific subject matter.

[0017] Figure 2 illustrates a detailed view of the system 200 for estimating a deficit in knowledge required for completing a task presented to at least one user and further configured for remedying the estimated deficit in knowledge required for completing the task, based on at least one user’s as well as all cumulative users’ data of interaction.

[0018] Figure 3 illustrates an example flow chart depicting the flow of the user interaction with the system of Figure 2. [0019] Figure 4 illustrates a flowchart depicting the logic flow in the server, the ML based concepts processor, and the user system.

[0020] Figure 5 illustrates the use of ML based Concepts Processor of Figure 2 with respect to one or more of problem solving interactions for a plurality of users, with the system of Figure 2.

[0021] Figure 6 illustrates a flowchart of a method for estimating a deficit in knowledge required for completing a task presented to at least one user.

[0022] Figure 7 illustrates the flowchart of a method for remedying the estimated deficit in knowledge required for completing the task, by the at least one user.

[0023] Figure 8 is a block diagram of a computing device utilized for implementing the system (200) of Figure 1.

[0024] Further, skilled artisans will appreciate that elements in the figures are illustrated for simplicity and may not have necessarily been drawn to scale. Furthermore, in terms of the construction of the device, one or more components of the device may have been represented in the figures by conventional symbols, and the figures may show only those specific details that are pertinent to understanding the embodiments of the present disclosure so as not to obscure the figures with details that will be readily apparent to those of ordinary skill in the art having the benefit of the description herein.

DETAILED DESCRIPTION

[0025] For the purpose of promoting an understanding of the principles of the invention, reference will now be made to the embodiments illustrated in the figures and specific language will be used to describe the same. It will nevertheless be understood that no limitation of the scope of the invention is thereby intended, such alterations and further modifications in the illustrated system, and such further applications of the principles of the invention as illustrated therein being contemplated as would normally occur to one skilled in the art to which the invention relates.

[0026] It will be understood by those skilled in the art that the foregoing general description and the following detailed description are exemplary and explanatory of the invention and are not intended to be restrictive thereof. [0027] The terms "comprise", "comprising", or any other variations thereof, are intended to cover a non-exclusive inclusion such that a process or method that comprises a list of steps does not comprise only those steps but may comprise other steps not expressly listed or inherent to such a process or a method. Similarly, one or more devices or sub-systems or elements or structures or components proceeded by "comprises... a" does not, without more constraints, preclude the existence of other devices or other sub-systems or other elements or other structures or other components or additional devices or additional sub-systems or additional elements or additional structures or additional components. Appearances of the phrase “in an embodiment”, “in another embodiment” and similar language throughout this specification may, but do not necessarily, all refer to the same embodiment.

[0028] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. The system, methods, and examples provided herein are illustrative only and not intended to be limiting.

[0029] In addition to the illustrative aspects, exemplary embodiments, and features described above, further aspects, exemplaiy embodiments of the present disclosure will become apparent by reference to the drawings and the following detailed description.

[0030] In some embodiments, the word ‘user’, ‘learner’, ‘student’, and ‘individual’ used in the description may reflect the same meaning and may be used interchangeably. In some embodiments, the word ‘knowledge gap’, ‘learning gap’, ‘deficit in knowledge’ and ‘lack of knowledge’ used in the description may reflect the same meaning and may be used interchangeably. In some embodiments, the word ‘problems’, ‘series of questions’, ‘task’ and ‘one or more questions’ used in the description may reflect the same meaning and may be used interchangeably. In some embodiments, the word ‘filling the knowledge gap’ is referred to as as “remedying the estimated deficit in knowledge required for completing the task, for completing a task presented to at least one user.

[0031] It is to be noted that in some embodiments disclosed herein, completing a task may mean solving a problem, answering a question, doing an assignment, answering a multiple choice question, which may itself require solving a mathematical problem, a problem in physics that requires both a knowledge of concepts in physics and solving a mathematical problem using the concepts, using knowledge of a subject and using logic to arrive at the solution and so on. It is to be noted that, the scope of the term ‘solving a problem’ is not limited to only mathematics and physics as stated above, but the same concepts of mathematics and physics may be implemented for computer programming or any other subject matter. It may not only include the physical sciences but also natural sciences, economics, social sciences, and so on. The disclosed system and method can also be implemented for any academic subject. The tasks related to computer programming may be assigned to the user by just changing the content and the corresponding concepts of any other subject matter.

[0032] The term knowledge and skill used herein is explained as follows. It is to be noted that, the term knowledge as used herein can be defined as the concepts learned by attending classroom lectures, watching videos on an online learning platform, and so on. Computer programming may be considered as an example of a subject matter other than mathematics or physics. Computer programming may be used for structural engineering' in Civil Engineering or Network Theory in Electrical Engineering. The system 200 as disclosed herein with reference to Figure 2 is configured for improving the problem-solving skills of the user of the system. The body of knowledge is represented by a set of concepts corresponding to each subtopic, and topic of a subject matter. The term ‘skill’ means, specifically in this case, the problem-solving skill that is gradually built by the user by using, attempting to solve, or solving a series of curated problems provided to the user by the disclosed system. The system as disclosed herein estimates the gaps in the user’s knowledge precisely, pertaining to the problem they are solving. The knowledge gap is filled by using the method as disclosed in Figure 7, so that the user, even with a poor academic background, will consolidate his knowledge along with his problem solving skills. The term ‘skill’ as used herein can be defined specifically as 'problem-solving skill’, the central subject of the current disclosure. Imbibing knowledge may not necessarily give the user the skills to solve complex problems which is the central theme here. In one example, the user gains the knowledge of a concept which are required to solve the problem presented to the user. The disclosed system not only fills the knowledge gaps but also improves their problem-solving skills which are important to do well in the evaluation tests and examinations, which are also time bound. The skill is all about understanding the problem at hand and then applying a combination of concepts in the proper sequence within the time given to solve the problem.

[0033] Embodiments of the present disclosure discloses modules configured for estimating a deficit in knowledge required for completing a task, and even to build problem solving knowledge base from ground-up or the level(s) where the user requires to build their knowledge from. Briefly, according to an exemplary embodiment, disclosed herein is a system and method for building knowledge required for completing a task in a specific subject matter. The system as disclosed herein is a “Machine Learning based learning system” for building knowledge for solving a problem or completing a task for a specific subject matter associated with an academic subject.

[0034] The term ‘building knowledge’ may be defined as follows in the present context. The system gives a one or more tasks to the user. In other words, one or more questions to be answered. In this process, the system identifies the knowledge gaps using time series neural network and the knowledge gap is filled with the help of clustering of individual concepts in each sub-topic and topic, and group of concepts pertaining to each task. There are a finite number of concepts for each sub-topic of the subject matter. The entire process of identifying and filling the knowledge gaps for solving problems in the context of the present disclosure is called building knowledge.

[0035] The disclosed method and system further facilitate in estimating a deficit in knowledge required for completing a task presented to at least one user and filling of knowledge gaps of the user in completing the task, based on this user’ s as well as all cumulative users’ data of interaction with this disclosed learning system.

[0036] Briefly, according to an exemplary embodiment, disclosed herein is a method and system for building knowledge of a user in a specific subject matter. The disclosed method and system aid in identifying and filling the specific gap or gaps in knowledge required for solving a specific problem, specific to the learner at the time or at any instance or stage of solving the problem, based not only on the various parameters of the user’s interaction with the disclosed learning system, but also on the parameters of all users who have interacted with this learning system. Since this learning system is based on Machine Learning (ML) and Artificial intelligence (Al), it becomes better with every interaction of each user since the data captured is implemented to train and improve the learning model further. The filling of the knowledge gap or gaps is specific to the user’s competence at the time of solving the problem.

[0037] Embodiments of the present disclosure will be described below in detail with reference to the accompanying figures. [0038] FIG. 1 illustrates a high-level block diagram of a machine learning (ML) based system 100 configured for ‘building knowledge’ for concepts-based problem solving in a specific subject matter. Figure 1 depicts a high-level block diagram of the system 100, where a plurality of students/users are interacting using the user devices 101-B-N with the learning system 100 along with a number of tutors (for example a tutor using a user device 101-A) who are also on this system to interact with students using user devices 101-B-N through the systems.

[0039] The students’ and tutors’ interaction with the user devices 101-A, 101-B, . . . 101-N is communicated to web applications 103 using one or more data communication methods.

[0040] The frontend system 104 communicates this information to the backend system 105 where the concept processor is configured for executing the steps for building knowledge. The working of the concept processor is explained in detail below. This students and tutors’ interaction with the user devices 101-A to 101-N is stored in a database and used for determining the second data. The use of second data is explained in detail below with reference to Figure 2. The second data is the data collected in advance or collected historically, based on each interaction of student’s and tutors’ while completing each task of a plurality of tasks, associated with an academic subject, presented to each of the students and tutors. The second data can be defined as the data of interaction of universal set of users interacting with the system for all tasks cumulatively.

[0041] The backend system 105 also has bidirectional communication with two sets of databases. The first database 106 stores one or more of tasks comprising multiple choice questions (MCQs), videos, and assignments. The second database 115 stores the students’ details and parameters. The system 100 as disclosed herein includes a user device comprising a data capture device to record the interaction between the system 100 and the students using the user device 101 -A to 101-N, while the students are completing a task. Briefly, according to an exemplary embodiment, disclosed herein is a system and method for building knowledge of a user for solving a problem in a specific subject matter. The system as disclosed herein is a machine learning based learning system for building knowledge of a user required for solving a problem or completing a task in a specific subject matter associated with an academic subject. Examples of subject matter are Class X Mathematics: Trigonometry; Electrical Engineering: Network Theory. Examples of assignments are, but not limited to: Mathematics, Class X, CBSE, Trigonometry: “Show that (sin ⁴ x) ² - (cos ⁴ x) ² = (sin ² x - cos ² x)(l-2sin ² xcos ² x) Mathematics, Class XU, CBSE, Integral Calculus as shown below.

[0042] The term ‘building knowledge’ can be defined as follows. The system gives one or more one or more tasks to the user, in other words, one or more questions to be answered. In this process, the system identifies the knowledge gaps, using time series neural net, and the knowledge gap is filled with the help of clustering of individual concepts in each sub-topic and topic, and group of concepts pertaining to each task. There are finite number of concepts for each sub-topic of the subject matter. The entire process of estimating a deficit in knowledge required for completing a task presented to at least one user and filling of knowledge gaps of the user required for completing the task, is called as building knowledge.

[0043] A manner in which the system 100 is configured for estimating a deficit in knowledge required for completing a task presented to at least one user and further configured for filling of knowledge gaps of the user in completing the task, based on that user’s as well as all users’ cumulative data on interaction is described in detail further below.

[0044] Figure 2 illustrates a detailed view of the system 200 for estimating a deficit in knowledge required for completing a task presented to at least one user and further configured for remedying the estimated deficit in knowledge required for completing the task, based on at least one user’s as well as all users’ cumulative data of interaction.

[0045] It is to be noted that the system 200 as shown may be explained with respect to a single user, however, it should be noted that the present disclosure can be similarly applied to multiple users. Further, the user using the user device 201 may communicate with the system 200 using one or more user devices (exemplary user device 201) through a network (not shown) using a data communication device. Examples of the user device 201 include, but not limited to, a mobile phone, a computer, a tablet, a laptop, a palmtop, a handheld device, a telecommunication device, a personal digital assistant (PDA), and the like. Examples of the network include, but not limited to, a mobile communication network, a Local Area Network (LAN), a Wireless Local Area Network (WLAN), a Wide Area Network (WAN), internet, a Small Area Network (SAN), and the like.

[0046] In one aspect of the embodiment, the system 200 is configured for estimating a deficit in knowledge required for completing a task presented to at least one user. In another aspect of the embodiment, the system 200 is configured for remedying the estimated deficit in knowledge required for completing the task. Both aspects of the mentioned embodiments are explained in detail below.

[0047] The system 200 includes the user device 201 configured for collecting a first data based on an interaction of the at least one user with the user device 201 while completing the task presented to the at least one user on the user device 201. In one example, the task may include one or more problems to be solved or attempted by at least one user, presented on the user device 201. The first data is defined as the data of the user created by the system 200 while the user is submitting the completed tasks given to them (The term them here is the genderneutral singular in place of her or them and may be used throughout this disclosure in that sense) by the system 200.

[0048] The system 200 is configured to maximize the user’ s learning for the specific subj ect matter with the help of a user device 201 by displaying one or more one or more problems, collecting the user interaction from the data capture device 202 in the user device 201, and transmitting this captured data to the application servers 220. Using the data communication device 210, the user interactions (the first data) are collected by the application server 220 and stored in the database. The data communication device 210 of the user device 201 is configured for transmitting the first data to the server 220 for analysing the first data based on a first set of parameters and a second set of parameters. The first set of parameters includes data associated with a submission of each completed task by the at least one user, wherein the data is related to a combination of concepts present in the task, level of difficulty in completing the task, time given to complete the task, time taken by the user to complete the task, number of concepts referred to or used by the user to complete the task, number of attempts by the user to complete the task, result associated with task solved. In one example, the level of difficulty in completing the task or time given to complete the task to the user may be based on decision or opinion of a subject matter expert in the specific academic subject or a consensus opinion of a group of experts and so on. For example, the expert may take into consideration the total number of concepts needed to solve a problem or completing a task as a measure of difficulty. [0049] The users’ parameters 215 which are the second set of parameters are also collected and sent to a data preprocessor 207. The second set of parameters are determined based on the data associated with each interaction of each user of a plurality of users while completing each task of the plurality of tasks, associated with the academic subject, presented to the each of the plurality of users. For example, the second set of parameters 215 are determined, based on all users’ cumulative data of interaction of those who have interacted with the learning system 200.

[0050] The modules of the system 200 collects, collates, and pre-processes the following first set of parameters associated with the submission of answers to each problem by the user such as problem ID, one or more concepts required to solve each problem, level of difficulty of each problem, time given to solve the problem, time taken by the user to solve the problem, number of concepts referred to by the user to solve this problem, level of help taken, number of attempts by the user to solve this problem, solution to the problem, and whether the problem was solved or not.

[0051] This pre-processed data is then fed to a Machine Learning Engine - ML based concept processor 225, in communication with the application server 220 which provides computation of the user’s competence, and accordingly fill the knowledge and skill gap for solving the current problem, which is then sent to the user’s feedback device 203. The ML based concepts processor is a combination of AI/ML technologies, designed to minimize the ‘Cognitive load’ for the user when learning and solving problems of a subject matter associated with an academic subject.

[0052] The application server 220 is configured for receiving the first data for training a model iteratively, using a second data. The second data is the data collected in advance or collected historically, based on each interaction of each user of a plurality of users while completing each task of a plurality of tasks, associated with an academic subject, presented to each of the plurality of users. The second data can be defined as the data of interaction of universal set of users interacting with the system for all tasks cumulatively. This second data is used for training the time-series neural network iteratively, whereas the first data, pertaining to a specific user using the user device 201 is used for inferring their knowledge gaps. This second data as well as the current user’s live data are directed to the application server 220 from where it goes to the data preprocessing 207 where the data is cleaned, correlated, the categorical data is encoded. [0053] The model is trained for determining a familiarity score for the at least one user indicative of familiarity of the user with the academic subject based on a learning history of the user. The model is trained for determining a competence score for the at least one user, based on analysis of the first data. The term ‘learning history’ may be explained with an example. Let us consider the term ‘learning history’ from the context of a chapter in mathematics such as Trigonometry. The user may be briefed with simple concepts such as basic trigonometric identities. The user may be assigned a simple task (questions) to be solved based on basic trigonometric identities. The various parameters, such as first set of parameters (as explained above, for example, level of difficulty in completing the task, time given to complete the task, time taken by the user to complete the task, number of concepts referred to or used by the user to complete the task), of this user will be recorded by the system as the first data. The ML based concept processor 225 of the system 200 is configured to identify the knowledge and the skill gaps of the user based on these initial interactions. Subsequently, more complex problems will be given to this user which may be a combination of basic trigonometric identities, factorizations concepts, simultaneous equations, or triangles. The other topics may belong to concepts of earlier chapters of the same year or any of the earlier classes or years. The time-series model neural network of the ML based concept processor 225, is then configured to derive the user's learning history based on the data collected while completing the assigned tasks.

[0054] In one embodiment, the model is trained by implementing a time-series neural network. For example, the time-series neural network model is trained on the data of multiple students’ solution submission. However, while estimating a user’s learning stage and their knowledge gap or gaps, only their data of interaction (i.e., first data) is used. The trained data is used for estimating the deficit in knowledge required for completing a task presented to the at least one user, using the output of the model and the analysed first data. The output of the model predicts the degree of knowledge gaps of the user for completing a task presented.

[0055] The time-series neural network gives inferences for each group of concepts pertaining to each task present to the user. It uses a scale of 1 to 5, commensurate with Bloom’s taxonomy levels (as known in the state of art) of 1 to 5. This level shows the degree of mastery of the user for a given group of concepts related to a specific task. The Gaps in knowledge are estimated in a scale of 1 to 5 by the machine learning model. In one example, the rating of 5 may be considered as the student is ready for exam. In another example, the rating of 1 may be considered as very poor. The model estimates in these scales since the teaching standard 'Blooms Taxonomy' is also having a pyramidal shape where the learning stages are stacked from 1 to 5.

[0056] The determination of familiarity is performed using a method involved which takes care of the familiarity of the user to solve a type of problem. This requires the data pertaining to learning history of a particular user. By using a time series neural network model, a user’s instantaneous competence in solving a particular problem is extrapolated by the model. Simultaneously the degree of gap or gaps in their knowledge to solve this problem is predicted. The input to this model is the series of first set of parameters when the user submits the answers to questions posed by the system. It is this model which identifies the knowledge gaps of the user in problem solving. The system 200 enables the user to clear their basic concepts as well as learn complex problem solving.

[0057] The ML based concepts processor 225 then outputs the identification of the gaps in the user’s knowledge and skill for solving the current problem and accordingly provides the user with specific intellectual scaffolding to fill in their knowledge and skill gap or gaps for the specific problem, thus not letting the user get frustrated over the unsolved problem. Traditionally, the intelligent scaffolding was done by giving hints to the user for solving the problem. This is a vague concept. The system 200 is configured for providing the list of concepts required by the user for completing a task comprising solving a problem, and out of that list, a sub-list required for solving the problem required, by this specific user is obtained based on his learning history and the clustering of group of concepts for each task.

[0058] In another aspect of the embodiment, for remedying the estimated deficit in knowledge required for completing the task, for a user to complete the task comprising a specific problem, a clustering module 224 configured for clustering of concepts at two levels is implemented. The filling of knowledge gap involves remedying the estimated deficit in knowledge required for completing the task, by the at least one user.

[0059] At a lower level, the clustering is done based on sub-topics and topics pertaining to the subject matter associated with an academic subject. At a higher level, the clustering is done on group of concepts involved with each task. The following paragraph explains the working of the clustering module 224 in detail. [0060] In one example and as mentioned above, a task includes one or more of questions. A question in a subject matter may have a combination of concepts required to solve it. The question may not only require concepts from various chapters of a class year but may encompass concepts from earlier class years. Various questions may require a huge number of these concepts to answer it.

[0061] The clustering module 224 is configured to collate an exhaustive body of questions in the subject matter associated with various academic subjects. It then encodes each group of concepts required for answering each question. The encoded concepts are then embedded, and this serves two purposes. The first is dimensional reduction so that the computations involved are drastically reduced. The second purpose is a sense of clustering of concepts belonging to topics and sub-topics. A topic that is part of a broader or more general topic (for example., a chapter) and each topic branches off into sub-topics. A topic example is trigonometry. The subtopics for trigonometry can be trigonometric ratios, trigonometric identities, applications of trigonometry. It is to be noted that this classification of topic and sub-topic may vary in detail and a person skilled in the art will be able to decide on the classification.

[0062] Each group of concepts (for each question) is now grouped based on one of the clustering algorithms which can handle multidimensional objects. In one example, k-clustering method known in the state of art may be implemented herein. It is to be noted that, other known methods of clustering can also be implemented, and the disclosed method is not limited to only k-clustering. These groups of concepts are clustered based on the similarity of the groups, and this is used for reinforcing the user’s problem solving capability (filling the knowledge gaps). In one example, Euclidean distance method may be used to measure the similarity between two texts based on the angle between their word vectors. It is to be noted that, other known methods of similarity can also be implemented and is not limited to only Euclidean distance method.

[0063] At a lower level, the clustering is done based on sub-topics and topics pertaining to the subject matter. In one embodiment, the clustering module 224 is configured for clustering of plurality of concepts based on sub-topics and topics pertaining to each task of the plurality of tasks associated with the academic subject presented to the each of the plurality of user.

[0064] At a higher level, the clustering is done on group of concepts involved with each task. In another embodiment, the clustering module 224 is configured for clustering of plurality of concepts based on group of concepts involved with each task of the plurality of tasks associated with the academic subject, presented to the each of the plurality of users.

[0065] The steps for clustering the plurality of concepts include collating a plurality of tasks presented to each of the plurality of users, obtaining, combining, and encoding each group of concepts involved with each task of the plurality of tasks associated with the academic subject presented to the each of the plurality of users; embedding the encoded group of concepts into a memory coupled to a concept processor 225 of the server. As known in the state of the art, the Word2vec is a method used for creating word embeddings and is used to obtain embedded group of concepts and the same has been used herein. Word embeddings is a combination of word tokenization and embeddings processes, since ML mathematical models can only process numbers and cannot process words. There are multiple algorithms for embeddings such as for example, Word2Vec, Glove, Embedding, and similar such algorithms known in the state of art. Embedding has two uses. The first one is dimensionality reduction, which translates into reduction in computations. In one example, the knowledge may be a combination of 500 concepts. Downstream in the ML model, it will require computations which are an exponential order of this number. By embedding this to say 50 for example, the computations are reduced by an order of magnitude. The second purpose of the embedding of concepts is to cluster the group of concepts for further processing. Embeddings make it easier to do machine learning on large inputs such as sparse vectors representing words.

[0066] The embedded group of concepts relating to each task are clustered using one or more clustering algorithms. This ensures that the core of similar tasks is clustered together in a multidimensional space. This ensures proper knowledge gap filling pertaining to various tasks where the competence of the user is lacking even to a small degree. The embedded encoded concepts stored in the memory are used for remedying the estimated deficit in knowledge required for completing the task, by the at least one user.

[0067] The steps described above are performed by the clustering module for clustering a plurality of concepts. The first step includes collating a plurality of tasks presented to each of the plurality of users.

[0068] At level 1, the clustering module is configured for obtaining, combining, and encoding concepts based on sub-topics and topics involved in carrying out each task of the plurality of tasks associated with the academic subject, presented to each of the plurality of users.

[0069] At level 2, the clustering module is configured for obtaining, combining, and encoding each group of concepts involved with each task of the plurality of tasks associated with the academic subject presented to the each of the plurality of users. Each group of concepts involved in solving each task of the plurality of tasks are clustered based on a predetermined similarity score.

[0070] The clustering module is configured for then embedding the encoded group of concepts into a memory coupled to the 225 in the server and using the embedded encoded concepts stored in the memory for filling the estimated lack of knowledge required for completing the task, by the at least one user. The embedded encoded concepts may include a specific concepts list computed by the clustering module 224 and provided to the user to minimize the number of concepts presented to the user so as to enable the user to solve the problem being attempted by them.

[0071] In an embodiment of the present disclosure, the Machine Learning engine - ML based concepts processor 208 improves its own performance when more number of users interact with this disclosed system, thereby providing improved assistance to the users, as time goes by.

[0072] Thus, the system 200 strives to fill all knowledge and skills gap or gaps of the user for solving a problem in the subject matter, with one or more of interactions with the user to lead them to mastery of the subject matter, irrespective of their foundational knowledge and skill spanning previous classes or years. The present disclosure provides a system 200, so that the user masters a particular subject matter in the best possible manner by providing them one or more of assignments, to assist them immediately if they cannot complete the given assignment, by providing the knowledge and skill missing in them to complete it. These specific knowledge items provided by the disclosed system go beyond the specific class or grade or chapter of the subject matter of the user, thereby ensuring that the fundamental knowledge missing in the user is also exhaustively covered. In this regard, the user’s foundational gaps do not matter anymore.

[0073] The present disclosure provides a method and system to identify the gap or gaps in the knowledge and skill required while attempting to solve a particular problem. After capturing one or more of problem-solving interactions with this disclosed system, the system comprehensively and precisely determines the gap or gaps in the user’s knowledge of subject matter. The present system, based on the user’s competence, and the user’s knowledge gap or gaps for a particular problem, provides the user with the exact concepts missing with them to solve it.

[0074] The disclosed system is based on machine learning and has a mechanism for improving its own performance in assisting users to learn better with every interaction of all users interacting with the present system.

[0075] The system 200 provides a self-optimizing system which makes users learn a subj ect matter faster and better by identifying the gap or gaps in their knowledge and skills required to solve problems and by utilizing the method of concepts. The system 200 helps in consolidating the user’s mastery of various subject matters at their current academic level, class, or year. The components and subsystems of the system 200 are explained in detail below.

[0076] Figure 3 illustrates an exemplary flow chart depicting the flow 300 of the user interaction with the system 200 of Figure 2. At step 342, the user is presented with a task for completion. In one example, the task may include a series of videos, for him to understand the various sub-topics with a chapter of the subject matter. At step 344, the user is then presented with a series of MCQs (Multiple Choice Questions) so that his level of understanding of the sub-topic is tested. Here, the distinction must be emphasized that the user has only understood the concepts and he does not have the knowledge and skills to solve the problem related to the sub-topic of the chapters. According to the flowchart 200, the user will now be presented, at step 346, with simple problems related to the sub-topics only. If they do not have adequate foundational knowledge to solve this problem, they will be assisted by the ML based concept processor at step 354, which will identify his knowledge and skill gap or gaps pertaining only to this assignment and present these concepts to enable the user to solve it. Once all sub-topics or assignments of the chapter are covered, the user will be presented with complex problems at step 362. Here again, if the user is not able to solve a problem, the ML based concept processor 225, at step 354, identifies the user’s knowledge and skill gaps, and then present these identified concept(s) to the user through the feedback device 203 of the user device (as shown in Figure 2). In case of complex problems, the list of concepts required for a particular problem will be more elaborate as compared to the simple problems in a sub-topic. The terms ‘assignments’ and ‘problems’ are used interchangeably in the disclosure. [0077] Once the user has successfully solved complex problems which span not only all sub-topics of the chapter for a subject matter, but it may also contain concepts from different chapters, and also may contain concepts from previous classes, the user will be deemed competent, at step 370, on this chapter, including dependent concepts from previous classes or grades.

[0078] Figure 4 illustrates a flowchart depicting the logic flow 400 in the server, the ML based concepts processor, and the user system. FIG. 4 illustrates the flowchart 400 depicting the logic flow in the application server, the ML based Concepts Processor, and the user device of Figure 2. The flowchart 400 shows that with the help of concepts as building blocks, the process of assisting the user in solving problems, from simple to complex in a subject matter. The flowchart 400 also shows that this system 200 computes the competency level of the user at any instant. The flowchart also depicts the system and method of identifying the knowledge and skill gaps, in terms of a list of concepts, to solve a particular problem. This helps user to become familiar with the problem and its solution thereby enabling him to constructively apply his mental effort to solve it smoothly, without which the user may face frustration.

[0079] The flowchart starts at step 472. At step 474, the user interacts with the system 200 for completing the task comprising solving problems presented to the user on the user device. At step 476, the first data based on the user interaction is collected and transmitted to the application servers. The first data is defined as the data of the user created by the system 200 while the user is submitting the completed tasks given to them by the system 200. The server is configured to process (at step 478) the inputs (the first data) from the user and the second data. The application server is configured for receiving the first data for training a model iteratively, using a second data. The second data is the data collected in advance or collected historically, based on each interaction of each user of a plurality of users while completing each task of a plurality of tasks, associated with an academic subject, presented to each of the plurality of users. The second data can be defined as the data of interaction of universal set of users interacting with the system for all tasks cumulatively. This second data is used for training the time-series neural network iteratively, whereas the first data, pertaining to a specific user using the user device is used for inferring their knowledge gaps. This second data as well as the current user’s live data are directed to the application server from where it goes to the data preprocessing where the data is cleaned, correlated, the categorical data is encoded. At step 480, the user’s competency is analysed at the current instance. The user’s competency is determined using the competence score for the user, based on analysis of the first data. The model is trained for determining a familiarity score for the at least one user indicative of familiarity of the user with the academic subject based on a learning history of the user. The model is trained for determining a competence score for the at least one user, based on analysis of the first data and estimating the deficit in knowledge required for completing a task presented to the at least one user, using the output of the model and the first data.

[0080] At step 482, the concepts related to a particular problem are corelated. At step 484, the gaps in the user’s knowledge and skill for solving the current problem is identified. For corelating the concepts, and for remedying the estimated deficit in knowledge required for completing the task, for a user to complete the task comprising a specific problem, a clustering module as shown in Figure 2 is configured for clustering of concepts at two levels. The filling of knowledge gap involves remedying the estimated deficit in knowledge required for completing the task, by the at least one user. At a lower level, the clustering is done based on sub-topics and topics pertaining to the subject matter associated with an academic subject. At a higher level, the clustering is done on group of concepts involved with each task. At step 486, the gaps identified are communicated to the user using the feedback device 203.

[0081] Figure 5 illustrates the use of ML based concepts processor of Figure 2 with respect to one or more of problem solving interactions for a plurality of users, with the system of Figure 2. The system generated questions 501 -A to 501-N are presented to the plurality of users (shown as Question 501 -A to 501-N). The corresponding user response data 502-A to 502-N of the plurality of users are fed to the ML Concepts Processor 525 which then evaluates user capability 504 and based on the user competency, the ML concepts processor 525 identifies and fills the user’s knowledge and skill gap, in terms of the relevant concepts, which are communicated to the plurality of users using feedback device 203, as at step 506.

[0082] Figure 6 illustrates a flowchart of a method for estimating a deficit in knowledge required for completing a task presented to the at least one user. Figure 6 may be described from the perspective of a processor that is configured to execute computer-readable instructions to carry out the functions of the modules (described shown in Figure 2) of the system 200. In particular, the steps as described in Figure 6 may be executed for estimating a deficit in knowledge required for completing a task presented to at least one user. Each step is described in detail below. [0083] At step 625, a first data is collected and transmitted using a user device, based on an interaction of the at least one user with the user device while completing the task presented to the at least one user on the user device. At step 626, the first data is analysed based on a first set of parameters and a second set of parameters. At step 628, the first data and a second data is received by the server. The model is trained iteratively, using a second data, the second data being the data based on each interaction of each user of a plurality of users while completing each task of a plurality of tasks, associated with an academic subject, presented to each of the plurality of users. At step 630, a familiarity score is determined for the at least one user, indicative of a familiarity of the user with the academic subject based on a learning history of the user. At step 632, a competence score is determined for the at least one user, based on analysis of the first data. At step 634, a deficit in knowledge required for completing a task presented to at least one user is estimated.

[0084] Figure 7 illustrates the flowchart of a method for remedying the estimated deficit in knowledge required for completing the task, by the at least one user. Figure 7 may be described from the perspective of a processor that is configured to execute computer-readable instructions to carry out the functions of the modules of Figure 2 of the system 200. In particular, the steps as described in Figure 7 may be executed for remedying the estimated deficit in knowledge required for completing the task. Each step is described in detail below.

[0085] At step 736, a plurality of tasks is collated and presented to each of the plurality of users. At step 738, obtaining, combining, and encoding each group of concepts involved with each task of the plurality of tasks associated with the academic subject presented to each of the plurality of users is carried out. At step 740, obtaining, combining, and encoding concepts based on sub-topics and topics involved with each task of the plurality of tasks associated with the academic subject, presented to the each of the plurality of users is carried out. At step 742, embedding the encoded group of concepts into a memory coupled to a concept processor 225 of the server is carried out. At step 744, using the embedded encoded concepts stored in the memory for remedying the estimated deficit in knowledge required for completing the task, by the at least one user is carried out.

[0086] Figure 8 is a block diagram of a computing device utilized for implementing the system 200 of Figure 2. The modules of the system 200 described herein are implemented in the computing devices. The computing device 800 comprises one or more processor 802, one or more computer-readable RAMs 804 and one or more computer-readable ROMs 806, all being interconnected to one or more buses 808. Further, the computing device 800 includes a storage device 810 that may be used to execute operating systems 820 and modules existing in the system 100. The various modules of the system 200 can be stored in a storage device 810. Both the operating system and the modules existing in the system 100 and server 110 are executed by processor 802 via one or more respective RAMs 804 which typically include cache memory.

[0087] Examples of storage devices 810 include semiconductor storage devices such as ROM 806, EPROM, flash memory, or any other computer-readable tangible storage device 810 that can store a computer program and digital information. Computing device also includes R/W drive or interface 814 to read from and write to one or more portable computer-readable tangible storage devices 828 such as a CD-ROM, DVD, and memory stick or semiconductor storage device. Further, network adapters or interfaces 812 such as a TCP/IP adapter cards, wireless WI-FI interface cards, or 3G or 4G wireless interface cards or other wired or wireless communication links are also included in computing device 800. In one embodiment, the modules existing in the system 100 can be downloaded from an external computer via a network for example, the Internet, a local area network or other, wide area network and network adapter or interface 812. Computing device 800 further includes device drivers 818 to interface with input and output devices. The input and output devices can include a computer display monitor 818, a keyboard 824, a keypad, a touch screen, a computer mouse 826, and/or some other suitable input device.

[0088] While specific language has been used to describe the disclosure, any limitations arising on account of the same are not intended. As would be apparent to a person skilled in the art, various working modifications may be made to the method in order to implement the inventive concept as taught herein.

[0089] The figures and the foregoing description give examples of embodiments. Those skilled in the art will appreciate that one or more of the described elements may well be combined into a single functional element. Alternatively, certain elements may be split into multiple functional elements. Elements from one embodiment may be added to another embodiment. For example, orders of processes described herein may be changed and are not limited to the manner described herein. Moreover, the actions of any flow diagram need not be implemented in the order shown; nor do all of the acts necessarily need to be performed. Also, those acts that are not dependent on other acts may be performed in parallel with the other acts. The scope of embodiments is by no means limited by these specific examples. Numerous variations, whether explicitly given in the specification or not, such as differences in structure, dimension, and use of material, are possible. The scope of embodiments is at least as broad as given by the following claims.