CN104657567A - Student learning capacity evaluating method and system - Google Patents

Abstract

The invention is improvement of a student learning capacity evaluating method and a system. The system is compared with the 'psychology student file system', with user-friendly, simple and efficient problem does not repeat, can interpret evaluation results provide other advantages. The expert knowledge on the collected herein by reference for the evaluation of the Intel Education gauge formulation method, first the factors associated with high school students receiving learning technology as widely as possible through the packet, then aggregated portfolio, the conclusion that weight by Ranking process. Practice has proved that this method is efficient. In the data analysis phase, the invention introduces the world today recognized statistical class software SPSS, for correlation ANOVA every learning between technical factors and issues analysis, in addition to those with little relationship between sub-factors, so that the system becomes simple and effective.

Description

A kind of e-learning ability evaluating method and system

Technical field

A kind of e-learning ability evaluating method and system.

Background technology

Current, various circles of society are all in appealing, while alleviating the schoolwork burden of students in middle and primary schools, improve the learning quality of student, make learning process optimization, middle and primary schools' curriculum reform of a new round whole nation is carried out, and allows student's learning of learn, being the object of national curriculum reform, is also the requirement of social development.How to inquire into from the angle of psychological behavior the factor forming students in middle and primary schools' learning ability, prevent the generation of students in middle and primary schools' learning psychologies problem, research and solve effective educational countermeasures of students in middle and primary schools' learning psychologies problem and intervention antidote, testing, inquire into, research and develop the concrete learning art meeting students in middle and primary schools' learning psychologies, is the important topic that psychology and educational research person need conscientiously to research and solve.

" psychology of students in middle and primary schools' learning art is probed into " is one of Ministry of Education's Tenth Five-year plan emphasis problem, and problem is intended to inquire into the factor forming students in middle and primary schools' learning ability.At present, Application comparison widely students' learning ability test macro is " the students psychology archives economy " based on C/S developed by South China Normal University, and the volume of this system testing is large, and system maintenance is difficult, cannot make explanations to evaluation result.This patent uses the expert system of artificial intelligence (AI) to develop Middle school students ' learning capacity test system.

All the time, the research as aspects such as learning method, learning skill, learning strategies still can do nothing to help the overall process that we are familiar with Students ' Learning, more can do nothing to help student and starts with carry out the learning psychologies of construction oneself from changing oneself learning behavior.Angle research from psychological behavior is affected the factor of students in middle and primary schools' learning ability by this patent, by the mental mechanism that series of experiments omnibearing discussion students in middle and primary schools learning art is formed, make full use of pop psychology theoretical result and a set of efficient students in middle and primary schools' learning art training system of Information Technology Methods structure.

Summary of the invention

The function of this patent be by friendly with testee, alternatively simply talk with, analyze from metacognition technology, cognitive techniques, process technology, administrative skill four learning arts of aspect to testee, pointing out which aspect technology existing defects, providing foundation for improving targetedly.From I/O, system can propose some problems to testee in test process, and testee only need answer and be or no, and after being completed, the learning art of system to measured is evaluated, and provides evaluation reason.From security and privacy, the test of test macro to learning ability belongs to psychological test category, must meet the rule of psychological test and the principle of secrecy to testee's information.

Expert interviewing a kind ofly exchanges with aspectant the carrying out of expert the method obtaining knowledge.After expert group sets up, around this theme of Middle school students ' learning ability, according to the pattern of following table, adopt first indivedual summary, then concentrate the mode exchanged to obtain knowledge.Determine the problem of measurement in expert group after, organize the student and carry out survey.Survey will accomplish following 3 points: 1. the design of questionnaire is wanted to gather knowledge for problem; 2. adopt the method for science and pattern to conclude the data collected, comprehensively, reasoning from logic, and be processed into knowledge; 3. according to questionnaire, on purpose respondent is selected.Teach book series by consulting students in middle and primary schools' learning psychologies, understand the cognition of middle school student in learning activities, motivation, mood, behavior etc., centering student learning art is classified, and carries out the acquisition of knowledge more targetedly.。The data obtained due to survey are structurized data, are applicable to adopting the method for the data minings such as association analysis to find knowledge.

Survey is organized to be a rigorous and process for science, carrying out to make investigation work more in order, effectively, before survey, carrying out concentrating to student and instructing, help student understand questionnaire for problem and questionnaire involved by content, to ensure that survey can be carried out smoothly., specify problem, resolve into problem, put the project of survey in order.This time the theme of investigative action is Middle school students ' learning technical investigation and analytical evaluation, and this theme contains 169 investigation contents altogether, and be divided into four large class 17 groups, the distribution of problem is random., determine questionnaire object.Representative widely in order to make this time investigation have, in first middle part and high middle part, randomly draw 95 people by student number and participate in investigative action, with final whole city's general examination for students from various schools total score rank for achievement is according to institute, in class first 15 of rank think aristogenesis, the last 15 be backward pupil.The mode of filling in of questionnaire is taked to concentrate and is filled in., the basic comprising of questionnaire.Questionnaire is made up of the instruction (telling how surveyee fills in) of theme, questionnaire, points for attention and problem four part.

Embodiment

In order to the correlativity between each problem of analytic system and sub-factor, we are using the score of the average of all problems as sub-factor.Statistics software SPSS is selected to be data analysis software.Variance analysis, also known as analysis of variance or inspection, its objective is and infers that whether the population mean of two or more sets data is identical, check the difference of two or more sample average whether to have statistical significance.In order to determine the influence degree of the option value antithetical phrase factor of each problem, we select one-way analysis of variance.One-way analysis of variance is exactly whether the varying level testing some control variable causes significant difference and variation to observation variable.Because variance analysis has a stricter premise calls, totally should obey the identical normal distribution of variance namely under varying level, therefore variance analysis problem just converts the problem of the overall average under research varying level with or without significant difference to.The null hypothesis of one-way analysis of variance is HO: under the varying level of control variable, and each population mean is without significant difference.Namely the varying level of control variable does not produce significant impact to observation variable.

Namely the representation of knowledge become one group of computing machine acceptable data structure with the symbol of some agreements knowledge encoding about how describing things ~ group agreement.The method that the representation of knowledge is commonly used has State space representation and/or figure representation, semantic network representation, production rule representation, predicate logic representation, frame representation etc.TAL1.0 by with and/or figure representation and production rule representation represent the method for its knowledge.The general type of production is " former piece+consequent ".Former piece is exactly prerequisite, and consequent is conclusion or action, and former piece and consequent can form expression formula by logical operator AND, OR, NOT.Production describes a kind of corresponding relation (comprising cause-effect relationship and implication relation) between things.Its citation form is: " p-> " or " IFPTHEN Q ", and implication is if prerequisite P meets, then can release the operation of conclusion Q or execution Q defined.Wherein P is prerequisite or the former piece of production, and it depicts the condition precedent that this production could use, and forms by the logical groups of the fact is incompatible; Q is one group of conclusion or consequent, and it points out when prerequisite P meets, the operation that the conclusion that should release maybe should perform.

Tree search, exactly from start node, heuristically advances along the arc be connected with it, the process (also can back suction carry out) of searching destination node.So, after destination node finds, path also just have found.Conventional circle search strategy is divided into two large classes, and a class is exhaustive-search, and a class is heuristic search.Wherein exhaustive-search comprises BFS (Breadth First Search), depth-first search, bounded depth-first search, consistent cost search.Heuristic search is divided into local preferentially search, the overall situation is preferentially searched for and/or figure heuristic search, minimax search and A, B pruning search.Because the representation of knowledge employing in LTA1.0 and/or tree represent, and in solution procedure, do not have enlightening prompting, therefore, the search strategy in LTA1.0 is by the general search strategy adopted and/or set.

Solve-labeling procedure in search procedure and unsolvable labeling procedure are carried out all from bottom to top, namely by the solvability determination father node of child node, the solvability of grandparent node; By the unsolvability determination father node of child node, the unsolvability of grandparent node.And/or tree in, except end node and terminal node, the solvability of a node is decided by its child node completely.Depth-first search is exactly first only expand a child node all the time at every one deck of search tree, when constantly advancing until can not readvance (arrive leaf node or be subject to degree of depth restriction) to depth, just turn back to even higher level of node from present node, move on again along other direction.The search tree of this method from tree root one one formed gradually.

Knowledge base is made up of concept, the fact and regular three parts, indispensable.The concept of knowledge has been included in the fact, and in fact knowledge base comprises true Sum fanction two parts.Middle school students ' learning technical testing expert system comprises metacognition technology, cognitive techniques, process technology and administrative skill totally four aspects altogether.

Database reserves between some memory blocks in a computer, the fact deposited the fact of user's answer, the known fact with the fact depositing reflection system current state and obtained by reasoning, and that is namely derived in fact by own county magistrate assumes immediately, also as true.Native system is a test macro, and the database before test is the database be cleared, must by setting up dynamic data base with the question and answer of testee.The dynamic data base being set up system by question and answer can be reached by following statement.

The representation of knowledge has following two kinds of methods usually.One is, true and Data classification, be put in rule base, this expression is practically applicable to RBES.Based in the expert system of production rule, result is certainly the action of a production rule, and these production rules are by inputting data activation.Another kind method forms clause true and data, forms clause's knowledge base.This clause representation is practically applicable to the expert system of logic-based.This type systematic has a rule set that can be activated by input traffic.Native system is the system of a production, is applicable to adopting first method to represent knowledge.

Claims (4)

1. introduce when survey data analysis in the world today and generally acknowledge and popular comprehensive statistics analysis software package SPSS, carry out one-way analysis of variance (NOVA) between the sub-factor of learning art and problem, avoid the repetition of same problems.

2. in expressing for knowledge mode, use and/or the method for expressing such as tree, decision tree, table, and used the rapid translating between the various representation of knowledge of interactive expert system shell software simulating of being developed by Tom Cordon.

3. or the search of tree have employed the search strategy of depth-first.

4. being separated and backward reasoning mechanism of knowledge base (Repository) and inference machine.