FIELD OF INVENTION
The present invention generally relates to a system and a method for directed learning. More particularly, the invention relates to a system and a method for in-situ and directed learning through remote learning units.
BACKGROUND OF INVENTION
All current adaptive systems related to education are variations of computer adaptive tests. Computer adaptive tests use a bank of questions of varying difficulty levels to be served based on the student's responses. The main purpose of these adaptive tests is assessment, rather than improvement in the learning tools.
US Publication No. 20050026131 relates to providing a dynamic continual improvement educational environment. In particular, the present invention relates to dynamic systems and methods for gathering/tracking data, automatically adapting to information about an individual (e.g., the individual's pace of learning or other information), selectively determining the type and difficulty of content provided to an individual, selectively providing an exposure frequency for the content, and/or enabling rapid development and design modifications to the system and method for delivering learning units. The present system, in addition to providing differential learning to an individual based on his pace of learning, adapts according to the learner's specific responses. It tracks responses over time to identify learning gaps, misconceptions etc. existing in the learner and specifically targets those.

Learning is hindered sometimes by lack of conceptual understanding, sometimes by lack of practice and sometimes by both these. The present system can identify the exact nature of the learner's difficulties by tracking his responses and cater specifically to that.
US Patent No. RE39, 435 relates to systems and methods for personnel training and, more particularly, to supervised or self-administered computer-based training systems that incorporate a learner-constructed response based testing methodology for improved evaluation of knowledge acquisition. The disclosed system focuses more on improving evaluation to 'reduce guesswork'.
OBJECTS OF THE INVENTION
It is therefore an object of the present invention to propose an intelligent system and method for in-situ and directed learning from remote learning units.
Another object of the invention is to propose an intelligent system and method for in-situ and directed learning from remote learning units, which incorporates an improved learning module having data evolved from large-scale assessments of learners including an extensive classroom research.
A still another object of the present invention is to propose an intelligent system and method for in-situ and directed learning from remote learning units, which enables imparting a conceptual understanding of a study-field leading to expertise-building within the learner.

A further object of the present invention is to propose an intelligent system and method for in-situ and directed learning from remote learning units, which is enabled to operate from all the geographical locations including support-provisions in multiple languages.
A still further object of the present invention is to propose an intelligent system and method for in-situ and directed learning from remote learning units, which is operable on a continuous-basis under twin mode such as learning and evaluation.
SUMMARY OF INVENTION
Accordingly, there is provided A system for in-situ and directed learning through remote learning units, comprising an internet-based computer apparatus having at least one central processing unit, a plurality of memory devices, an input/output adapter for connecting several disk-storage units to a system bus, an user interface adapter for connecting to the system bus a plurality of input devices including a microphone and speaker, a display adapter for connecting the bus to an in-built display device, and a communication adapter for connecting the system to a communication network; and a software structure comprising at least one learning genome module having precisely defined conceptual webs and enabling mapping-out the entire learning context in several sequences eliminating therefrom the incorrect concepts so as allow a learner to acquire expertise on an academic curriculum and the conceptual understanding of a subject, the module being evolved based on a plurality of past performance data, and several acquired data through interviews of a large number of learners, the data being arranged in a specific hierarchy termed as clusters and each assigned a particular learning-levels; an

adaptive logic module having a level finder means enabled to determine admissibility of a learner at a starting level or a remedial level based on the responses acquired from the learner respectively in respect of an increasingly difficult question cluster or a remedial cluster; means for determining the difficulty value of each question including the learner's ability level to response in a dynamic and continuous fashion, the responses being analysed adapting a skill regression method in terms of learners errors which allow identification of weak and strong skills of the learner; and a learning tracking means enables a continuous evaluation of the learners through conducting pre-test and post-test on a regular interval with provisions of reports to the learners which allows a longitudinal tracking of weak areas including the improvements.
The present invention provides a system and a method for in-situ and directed learning through remote learning units which adapts an artificial intelligence module in an internet-based computer-apparatus for achieving learning including conceptual learning as well as procedural fluency (drill). The present system and method is an adaptive self learning system, for a learner to learn concepts and skills in academic subjects such as but not limited to Mathematics, Language arts, or the Sciences. The programme is delivered through a system that takes each learner along a unique path through a complex, but precisely defined set of concept maps that represent the relationship between the concepts. The concept maps are dynamic and iteratively modified based on both past data of over hundreds of thousands of students collected on a yearly basis, and dynamically based on current data as students use the system. Elaborate concept webs, large scale data and an intelligent system helps to identify the learning gaps, target misconceptions, and take the learner through an optimal learning path customized for any individual learner.

At the start, the system evaluates the learner's level and based on this, directs him or her to a particular starting point. The learner starts from fundamental concepts of this level, and learns by answering finely graded questions that represent specific learning goals. At every step, the learner's response determines whether he or she gets an explanation, practice or some other input to master that concept, or moves to a set of questions at a different location in the concept web. The system stores the learner's responses cumulatively, and uses this data to determine whether to take him or her to the next learning unit, when to offer a remedial, when to provide drill in a particular kind of task, when to throw a challenge and so on in order to provide an optimum learning trajectory. Hence, by combining the diagnostic and instructive capability of well-designed questions with powerful data-analysis tools and pedagogical intelligence in the form of the concept webs, the system assists the learner to progress entirely at his or her own pace and meet an unique set of learning challenges that he or she might face.
The present invention makes use of questions as primary learning tools for conceptual change. This is unique because the prior art exhibit a primary use of questions to assess and or as practice material in the form of exercises. The present system uses them as tools for building new concepts and addresses wrong concepts by actively getting students to solve problems by figuring out their underlying rules.
Another unique feature of the present system is that it is developed based on the performance data of over hundreds of thousands of students collected over time from various known diagnostic tests.

In contrast to prior art, the present invention adapts questions, not activities, to achieve learning. Each question typically is short and takes 20 seconds to 3 minutes at most to answer. Thus, unlike an activity, the goals and the sense of achievement is very clearly available to the student. Further, compared to its predecessor, each question is of a level of difficulty only infinitesimally higher. This too makes the challenge accessible to the student. At the same time, a learner who is clearly at a higher level (determined by the learner's performance over recent questions, as well as the time taken to respond to questions) the system 'jumps' levels to provide a question of greater difficulty, and in certain cases, specially designated 'challenge questions'. Yet, even these challenge questions are short and thus have a very clear goal. Thus, the disclosed system is entirely question-based, where the learner learns entirely through answering questions, with remedial feedback when necessary.
The present invention is a process for improving learning, making sure that the learner has gained an in-depth understanding of a concept, through a battery of questions given to him based on finding patterns of errors in his answers. The present inventive system has additional features which ensure this. The complex concept webs allow the learner not only to move across levels, topics or subjects with the aim of bringing about better understanding of concepts, and improved skill levels.
BRIEF DESCRIPTION OF THE ACCOMPANYING DRAWINGS
Figure 1 - Shows a pictorial view of the system of invention operably connectable to several nodes via internet, intranet, LAN.

Figure 2 - is a block diagram of the system architecture in accordance with a preferred embodiment of the invention.
Figure 3 - shows a flow-chart depicting the operational method of the system according to the invention.
Figure 4 - shows the steps performed by the Learning Genome module of the inventive system.
Figure 5 - shows the steps performed by the Adaptive logic module of the inventive system.
Figure 6 - shows a cluster arrangement depicting reflection of performance of a student on a particular question.
Figure 7 - depicts a process adaptable by the inventive system to improve the quality of learning service provided.
DETAIL DESCRIPTION OF THE INVENTION
The system as shown in figure - 1 constitutes a specifically designed hardware device that uses an embedded adaptive programme for learning a plurality of academic curriculum. The user has to connect this device to the internet to start working on it.
The system comprises:
- a Built-in device for read-aloud audio,

- an On/off button (2),
- a Display screen for questions, reports and other details (3),
- a Button to directly access user account details which is updatable by the user (4),
- a Button to go directly to the Topic Selection page (5),
- a QWERTY keyboard can be used to type in any value (6),
- a Button to directly access previous reports in the Report Module (7),
- a Left/right and up/down navigation button (8), and
- Option buttons (A, B, C and D) used to select correct answer options for Multiple Choice Questions (9).
Figure -2 depicts a system architecture for implementation of the disclosed invention.
A preferred embodiment of the system in accordance with the present invention can be practiced in the context of a personal computer, for example, the device of Fig-1. FIG. 2, illustrates a typical hardware configuration of the system in accordance with the preferred embodiment having a central processing unit (CPU), such as a microprocessor, and a number of other devices interconnected via a system bus (SB). The system shown in FIG-2 includes a Random Access Memory (RAM), Read Only Memory (ROM), an I/O adapter for connecting peripheral devices such as disk storage units (DSU) connected to a system bus (SB), a user interface adapter (UIA) for connecting a keyboard (KB), a mouse (M), a speaker (S), a microphone (MP), and/or other user interface devices to the bus (SB), a communication adapter (CA) for connecting the system to a communication network (e.g., a data processing network) and a display adapter

(DA) for connecting the bus (SB) to an in-built display device (DD). The system typically has resident thereon a known operating system. Those skilled in the art will appreciate that the present invention may also be implemented on platforms and operating systems other than those mentioned.
In other words, the system is configurable for example in the simplest way in the form of an internet-based device of Fig-1.
A preferred embodiment of the learning module can be written using high-level computer language and utilizes a programming methodology. A simulation engine in accordance with a preferred embodiment is based on a Visual Basic component developed to design and test feedback in relation to a spreadsheet.
These spreadsheet models are capable to simulate actual learning functions and become a 'task' that will be performed by a student The Simulation Engine accepts simulation inputs and calculates various outputs and notifies the system of the status of the simulation at a given time in order to obtain appropriate feedback.
As shown in figure -3, in the flow-chart, the Learning Genome is a detailed mapping of the learning in a topic into various subtopics, clusters and skills. This detailed mapping is done for all the subjects, and for all the topics for any curriculum.
A large collection of questions of gradually increasing difficulty levels are prepared by an experienced team.

The questions are mostly developed based on test data already available for thousands of students. New questions are tested in this manner. The exact sequencing and difficulty are decided based on this, and modified as students use the system.
In many cases, the large scale testing is followed by interviews of students to understand common misconceptions which are then addressed by specifically prepared questions.
Questions are made in different categories - basic, conceptual, practice-based, interactive and challenge as students must successfully be able to do all these types to achieve good learning.
A key innovation is an Adaptive Logic module (shown in fig-3) which determines the student's ability based on his response to a series of questions, identifies student strengths and weaknesses and determines a learning path.
An User Interface device presents different types of questions, explanations and facilities (e.g. for questions to be read out since reading is a major problem in rural Indian schools).
A Learning tracking means constitutes an extremely detailed record of student learning, and is maintained which can be called up even 1-2 years later to identify student learning patterns.

Learning Genome module (201)
As shown in figures 3 and 4 the Learning Genome (201) involves mapping out of academic learning in precisely defined conceptual webs. This is equivalent to mapping out the entire subset of specific facts, concepts and rules that students need to figure out in order to build expertise in that academic subject. Moreover, it includes several sequences in which these units should be learnt for an optimal learning experience. This mapping accounts not just 'learning from scratch', but also misconceptions and incorrect ideas that may have to be unlearnt or overturned. So it's a student-centered conceptual web and one that is increasingly based on data from students.
Detailed inputs from research are taken to create the complex concept webs, which allow movement not only within and across topics in a subject, but also across subjects. It is evolved from the performance data of hundreds of thousands of students in ASSET (201a), a leading diagnostic test, comprising multiple choice assessment items, and also from interviews (201 b) and first hand research into student learning. The data from large-scale assessments (201a) helps in identifying patterns of learning in students, pinpointing common misconceptions or areas of weakness etc. The student interviews (201b) and studies help in understanding the root causes of these weaknesses, misconceptions etc. The learning genome (201) is a complex web of concepts, sub-concepts and clusters (units of learning), arranged in hierarchical order, created based on this data. This hierarchy (201c) can be customized based on requirement of curriculum of different school education boards. Based on research and pedagogic rules, clusters are assigned multiple (201d) class levels. So if one cluster has been taught in a particular class and same has to revised in

the immediately upper class, that cluster will be assigned level I and II. The table below displays the learning genome prepared for example, on the topic "Algebra - basic concepts algebraic expressions and equations".
Algebra - basic concepts, algebraic expressions and equations

(Table Removed)

Furthermore, the content is organized in vertical strands, to enable movement across levels, to enable self paced learning. That is, the structure will allow the system to move a student of say, grade 7, to the grade 5 level concepts or material if need be.
Cluster Description:
A cluster is a collection of questions to teach a small sub-concept, e.g. concept of area or use of 'a' or 'an'. A cluster would normally have 30-50 questions. Similar questions are assigned same sub-difficulty level and there is minute increment of difficulty at each sub-difficulty level. Each cluster has a sequence (flow number) based on pedagogic rules.
Cluster classification:
1. Normal Cluster: Normal cluster is the cluster which is designed to teach a specific learning unit as per the sequence.
2. Remedial cluster: Remedial clusters are those clusters which are designed to address specific misconception or problem areas identified in students.
Adaptive Logic module (206)
As shown in figure -5, the Adaptive Logic module (206) is one of the key feature of the invention. Its functioning is described below:
By asking a set of questions chosen from the low, medium and high difficulty levels of each cluster, a Level Finder (206a) determines the starting level of each

student in each topic. Thus even if a student in class 7 is unaware of fundamental concepts of class 4, say, the Level Finder (206a) detects this and starts the student out at that level.
If a student understands the concept and able to answer questions correctly, he is taken to subsequent Questions clusters in the flow. But if he is not understanding the concept and not able to clear the Question cluster (206b), he is taken to remedial cluster (206b1).
Based on the student's performance, the Adaptive Logic (206) determines both his level and the type of practice he needs. Thus a beginning student (on a particular topic) will be given different question type (206c) for example, basic (introductory) questions, conceptual questions. Interwoven with these will be practice questions for drill. To reinforce learning, many modules have interactive questions. Finally a student is given challenge questions on that topic.
The Adaptive logic (206) dynamically updates (206d) both question difficulties and student abilities based on every question the student answer. In principle, if a learner of low ability answers a high difficulty question correctly, the questions difficulty will reduce marginally and the child's ability will increase marginally. Thus the system is self-correcting - errors in initial estimates of student ability or question difficulty level.
Item Response Theory (206e) is used widely in the field of Psychometrics used to analyze large scale performance data especially on tests, etc. It is also used by computer adaptive tests. Item Response Theory (206e) provides a more accurate and sophisticated tracking of student ability (which is done at the level

of each topic) and also allows the system to estimate the extent of guesswork, etc.
Every question is tagged with a number of skills (206c) (drawn from a large but finite list) with different weights. On completion of a cluster, a regression is run on all the questions completed by the student to determine strength areas and weakness areas. Questions tagged with the weakness area are given higher weightage in subsequent questions to provide the student practice.
Even after a module is completed successfully, the Adaptive Logic means (206) recognizes that the student may forget over time. At pre-defined periods after a topic is completed, a set of questions from the topic are asked again and if not answered satisfactorily, the system make the student to revise (206f) the topic again.
Adaptive logic (206) works on a huge bank of finely graded questions to pick out questions of different difficultly and sub difficulty levels. This is done with the help of the level finder module which records the response of each individual kid and based on the response type, picks up questions focusing on detailed concepts and sub concepts in a hierarchical pattern from cluster to subtopic to topic.
Here is an example for a subject such as 'Mathematics':
The Adaptive Logic module (206) consists of three pasub-modules:
1. Core module

2. Revision module
3. Timed test module
4. Games
The Core module consists of different clusters for various subtopics and topics. Each cluster consists of questions of a specific concept which are mapped at class level according to the difficulty level. The user starts off from the cluster of the lowest difficultly level for a given class. It picks questions from this cluster and works in a slider pattern, moving up and down based upon the responses-right or wrong at varying sub difficulty levels.
If the user is consistently getting the questions wrong, he is directed to a remedial cluster. Each cluster may act as remedial cluster depending on a child's specific concept being tested that needs scrutinized monitoring. When the user clears the remedial cluster he returns back to the level at which he was getting all responses consistently wrong. This allows the user to re-test on the same concept after clearing a lower level concept and move up on the slider.
When the user completes the cluster successfully, he is directed to the next cluster of a higher difficulty level. After every response, it gives a detailed explanation of the correct answer to help the user understand where he went wrong and thereby initiating the learning process. The responses given by the user is stored for analyzing the learning pattern. It remembers each individual user and starts him off from the same level at which left off.
Present System provides "Revision test" module at specified intervals by providing revision questions from different topics that need to be revised in that

month. The user's performance data is used to derive how far the learning has been achieved and at what level the user needs more brushing off of concepts.
The "Timed Test" module is designed to give the user a set of questions for practice in defined time duration, an essential feature in learning, like addition of two digit numbers, etc. The timed tests are mapped to a cluster and will be administered when the user finishes that cluster. These timed tests are repeated at regular intervals till the user is able to complete it successfully in the desired time interval.
A separate module has been developed for "Timed Test" which generates the questions dynamically based on the parameters defined by question developers. The parameters used in generating these questions are: 1. Format of the question, 2. Defining values passed to the variables used in the question, 3. number of questions to be given and 4. Time duration.
Games: Present invention teaches some concepts by asking questions in the form of games. It helps in better recalling of the concepts taught through some characters.
Item Response Theory (IRT), and Skill Regression Step (206f):
Item Response Theory is used to determine individual student's learning based on responses of a large group of students. The use of IRT to determine the difficulty value of each question and the ability level of each learner is a dynamic and continuous process. Theoretically as each question (of known difficulty level at that point of time) is given to each student (of known ability level), these

characteristics of both question and student change. Thus if a question initially marked as difficult is correctly answered by a number of 'weak' students consistently, its difficulty level will gradually reduce. Similarly, when strong students consistently get a question wrong, its difficulty value will increase.
Thus, the present system uses an Item-Response-Theory like framework to consistently adjust the question difficulty as well as student ability levels in a dynamic organization, based on live student data.
Figure 6 showing as to how a low student performance on a particular question type (SDL 4) is reflected when that learning unit is graphed. Corrections are made continually based on study of student data.
Another kind of error analysis is done using Skill Regression System where every question is linked to a number of skills. The students' responses to these questions help identify patterns of weak and strong skills. The methodology used is as follows:

(Table Removed)

Questions are linked to skills and responses analyzed to identify the underlying skill which may be leading to the error.
A regression equation run on the actual data provides the contribution of each of these skills to the student's performance in the set of questions. The accuracy of this exercise increases significantly as the number of questions increases. For example, the analysis may reveal that the contributors to the learner's score are as follows:
Skill Analysis of Student Code S2342 based on 30 Questions
(Table Removed)
The above is a dummy table since the number of skills in reality is very high -every subject topic and subtopic has specific skills related to understanding them, apart from common skills. The table above reveals three strength areas and three weakness areas. Of these, 'interpreting tabular information' is a key weakness (As every question is tagged, the system would likely give questions that strengthen understanding of this particular skill.)'
The present system also periodically performs the error analysis that identifies patterns of errors and misconceptions based on the data of an individual student

over time, and then provides him/her an opportunity to work on those specific identified errors. It also provides conceptual remedial as an opportunity for the student to improve conceptual understanding and minimize errors.
The Learner's error patterns are studied to identify the possible skill which lies at the root of the problem.
Because children often do not understand the concept of area they struggle to calculate the area of a shape which is a Simple combination of other shapes
(Formula Removed)

An example of how data from the diagnostic test helps to identify and analyze error patterns on an ongoing basis, which the present system targets through a series of different types of questions.
Hence the present system has the ability to iteratively collect intelligence on a learner through responses to questions and use this intelligence to determine the optimal learning trajectory and rate for any learner through the conceptual webs. Given that each learner has his/her unique circumstances, the system can figure

out how much and what route he/she must take in order to become an expert -based on storing the student's previous responses and on the basis of the known responses of previous students. So, the system is capable of understanding its current user and also understands users in general by accumulating experience by updating the conceptual web. This is one of the most unique features of the system. That is to say, the system is not just 'adaptive' in the sense that it adapts to the user on shorter time-scales, but actually become capable of being adaptive in the sense of updating its conceptual web based on learner experience over a longer time-scale.
As shown in figure -7, a daily time limit and duration of each individual session has been defined to regularize the user so as to achieve maximum learning.
The present invention aims to bring about learning, or improvement in learning, purely by asking the learner questions. These questions could be multiple choice questions or open ended questions. Based on the objective that whether a question is intended for learning, explaining or assessment, various categories of questions are:
Basic Questions: These are the questions that give basic understanding/introduction of the topic/concept to be taught. These questions explanation the concept and cover at the most one smallest unit of skill at a time.
Conceptual Questions: These are the questions designed to test the understanding of students at concept level and involve application of 2 or more skills to solve the problem.

Practice Questions: These questions are designed for practice purpose. Each Timed Test is a set of questions of same concept which are to be solved by student in limited time period. So these kinds of questions help in building speed in answering the questions.
Interactive Questions: Interactive questions require student to do some activity or manipulate some objects on the computer system. Interactive questions have been introduced to make students conversant with practical things in day to day life.
Challenge Questions: Challenge questions are designed to test the ability of students to apply the concepts/skills learned to solve a more complex problem. Challenge questions required student to recall multiple concepts, do the calculations and then solve the problem in a step by step manner.
Dynamically System Generated Questions: These questions are generated dynamically based on the parameters defined by question developers. The parameters used in generating these questions are: 1. Format of the question, 2. Defining range of the values to be passed to the variables used in the question. This technology helps in overcoming the limitation of no. of questions related to a particular skill/concept.
Voiceover Questions: Voiceover questions have been designed keeping in mind the reading problem in students of rural areas. System reads out questions for students.

Students also get the facility to put their comments/remarks which are addressed by EI team.
Reward and Motivation: The present system has means for reward determination based on performance to motivate and encourage the user. The system gives 'SPARKIES' or 'SUPER SPARKIES' on answering a series of questions correctly. The system gives points as reward for higher classes.
Admin/Teacher Interface: Specific Interface has been designed for school admin person and teacher to monitor progress of students and classes. Teachers can activate/deactivate topics for students. Teacher can view the individual students' progress and find out where the student is getting stuck. Teachers can revise a particular concept in class with the help of teacher interface. Most common wrong answers facilitate teachers to identify the problem areas/misconceptions in the class.
Learning Tracking means (208)
(Table Removed)

The present system has a provision for conducting Pre-test and Post-test. Based on the results of these two tests, the system determines the improvement in learner's understanding of concepts. The user is provided with reports on the progress made by him after every session. These reports also provide him information related to topic progress made, percentage of correct responses in each topic attempted by him, time taken per question attempt. It allows individual learning mode and group level learning comparison mode (referred as School mode). In a school mode, a user's performance is compared across class level. Also problem areas at class level are highlighted.
The inventive system and method for directed learning uses questions adaptively (using the artificial intelligence of a computerized system) as primary learning tools for conceptual change. The present system and method is an adaptive self learning system, for a learner to learn concepts and skills in academic subjects such as but not limited to Mathematics, Language arts, or the Sciences. The programme is delivered through a system that takes each learner along a unique path through a complex, but precisely defined set of concept webs that represent the relationship between concepts. The concept maps are dynamic, and iteratively modified based on data of over hundreds of thousands of students on a yearly basis. Elaborate concept webs and large scale data help pinpoint and target common errors and misconceptions.
The disclosed invention provides a system that places the learner at the centre of the learning. The system is enabled to adapt atleast three different teaching approaches that can be used:

1. Directive: Such a system is teacher-centered where the learner is instructed
on the subject matter.
2. Participative: Such a system is also teacher-led, but involves more
interaction and discussion with the learner playing a more active role than in a
directive system.
3. Learner Driven: In such a system, the learner drives the learning process
based on his or her prior knowledge, learning style, etc.
The current invention is a Learner Driven system that allows every student to learn at his or her own pace. It is a self-driven, self-paced learning system within an overall curriculum controlled by the teacher, school or parents. The system, described below, is a combination of carefully designed questions on the subject matter, large-scale data on those questions, computer technology, an adaptive self-learning program, statistical methods and feedback processes. It has currently been applied to learn basic subjects like Mathematics, language and science and can be extended to other subjects also. It can be an important innovation to help children learn better, especially in areas where learning levels are low and access to good teachers is limited (for example, rural, remote areas in India and other countries.)

WE CLAIM
1. A system for in-situ and directed learning through remote learning units, comprising:
- an internet-based computer apparatus having at least one central processing unit, a plurality of memory devices, an input/output adapter for connecting several disk-storage units to a system bus, an user interface adapter for connecting to the system bus a plurality of input devices including a microphone and speaker, a display adapter for connecting the bus to an in-built display device, and a communication adapter for connecting the system to a communication network; and
- a software structure comprising at least one learning genome module having precisely defined conceptual webs and enabling mapping-out the entire learning context in several sequences eliminating therefrom the incorrect concepts so as allow a learner to acquire expertise on an academic curriculum, the module being evolved based on a plurality of past performance data, and several acquired data through interviews of a large number of learners, the data being arranged in a specific hierarchy termed as clusters and each assigned a particular learning-levels;
- an adaptive logic module having a level finder means enabled to determine admissibility of a learner at a starting level or a remedial level based on the responses acquired from the learner respectively in respect of an increasingly difficult question cluster or a remedial cluster;

- means for determining the difficulty value of each question including the learner's ability level to response in a dynamic and continuous fashion, the responses being analysed adapting a skill regression method in terms of learners errors which allow identification of weak and strong skills of the learner; and
- a learning tracking means enables a continuous evaluation of the learners through conducting pre-test and post-test on a regular interval with provisions of reports to the learners which allows a longitudinal tracking of weak areas including the improvements.
2. A system for in-situ and directed learning through remote learning units as substantially described and illustrated herein with reference to the accompanying drawings.