CN103886054B - Personalization recommendation system and method of network teaching resources - Google Patents
Personalization recommendation system and method of network teaching resources Download PDFInfo
- Publication number
- CN103886054B CN103886054B CN201410093793.6A CN201410093793A CN103886054B CN 103886054 B CN103886054 B CN 103886054B CN 201410093793 A CN201410093793 A CN 201410093793A CN 103886054 B CN103886054 B CN 103886054B
- Authority
- CN
- China
- Prior art keywords
- resource
- teacher
- data
- course
- association
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
- 238000000034 method Methods 0.000 title claims abstract description 92
- 238000012545 processing Methods 0.000 claims abstract description 10
- 230000006399 behavior Effects 0.000 claims description 102
- 238000004458 analytical method Methods 0.000 claims description 35
- 238000001914 filtration Methods 0.000 claims description 14
- 230000009471 action Effects 0.000 claims description 12
- 238000004364 calculation method Methods 0.000 claims description 12
- 230000003542 behavioural effect Effects 0.000 claims description 9
- 238000003066 decision tree Methods 0.000 claims description 9
- 239000000284 extract Substances 0.000 claims description 9
- 230000003068 static effect Effects 0.000 claims description 7
- 238000010276 construction Methods 0.000 claims description 5
- 230000013011 mating Effects 0.000 claims 2
- 230000003993 interaction Effects 0.000 abstract description 2
- 238000010586 diagram Methods 0.000 description 17
- 101100194363 Schizosaccharomyces pombe (strain 972 / ATCC 24843) res2 gene Proteins 0.000 description 10
- 230000008569 process Effects 0.000 description 10
- 238000002156 mixing Methods 0.000 description 8
- 230000002452 interceptive effect Effects 0.000 description 6
- 239000011159 matrix material Substances 0.000 description 5
- 239000000463 material Substances 0.000 description 4
- 238000005065 mining Methods 0.000 description 4
- 210000003756 cervix mucus Anatomy 0.000 description 3
- 230000008859 change Effects 0.000 description 3
- 238000013461 design Methods 0.000 description 3
- 230000004048 modification Effects 0.000 description 3
- 238000012986 modification Methods 0.000 description 3
- 238000011524 similarity measure Methods 0.000 description 3
- 206010046901 vaginal discharge Diseases 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000000354 decomposition reaction Methods 0.000 description 2
- 230000003203 everyday effect Effects 0.000 description 2
- 238000000605 extraction Methods 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 239000000203 mixture Substances 0.000 description 2
- 230000000007 visual effect Effects 0.000 description 2
- 241001269238 Data Species 0.000 description 1
- 206010034719 Personality change Diseases 0.000 description 1
- 101100194362 Schizosaccharomyces pombe (strain 972 / ATCC 24843) res1 gene Proteins 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 230000000295 complement effect Effects 0.000 description 1
- 230000009193 crawling Effects 0.000 description 1
- 238000012217 deletion Methods 0.000 description 1
- 230000037430 deletion Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 238000011068 loading method Methods 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000012216 screening Methods 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
- 230000007306 turnover Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9535—Search customisation based on user profiles and personalisation
Landscapes
- Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention discloses a recommendation system and method of network teaching resources. The system comprises a data establishing module, an off-line data processing module and an on-line recommendation module. The data establishing module establishes teacher behavior data, teacher model data, a course model data and resource model data. The off-line data processing module is used for initializing and adjusting course model data and resource model data, teacher behavior data are used for deducing teacher identities, the degree of association between the resources is computed according to the teacher behavior data, the similarity between the resources is computed according to the resource model data, and the degree of association between the resources and courses is computed according to the resource model data and the course model data. The on-line recommendation module describes on-line recommendation resources through the association degree between the resources, the similarity between the resources, the association degree between the courses and the resources and the dynamic states of teachers, and the on-line recommendation module transmits teacher behavior data to the teacher behavior data of the data establishing module according to the feedback recommendation resource labels of the teachers on the recommendation resources through UI interaction.
Description
Technical field
The present invention relates to computer internet technical field, in terms of relating in particular to a kind of resource in networked teaching
Personalized recommendation system and its implementation.
Background technology
With the emergence of E-Learning, network teaching resource is also being increased in the way of swift and violent, and religion is given in the overload of information
Learn organizer and learner brings many challenges, they have to expend considerable time and effort, and just can filter out and meet
The teaching resource of oneself demand, therefore, the commending system that yields unusually brilliant results in commercial field, also begin to gradually be applied to education neck
Domain, it utilizes the historical behavior data of user, carries out personalized calculating, finds user interest point, and guiding user gradually finds to need
Ask information or resource, this improves the working and learning efficiency of user to a great extent.
Currently popular proposed algorithm have collaborative filtering recommending (Collaborative filtering, abbreviation CF),
Content-based recommendation (Content-based recommendation, abbreviation CB), Knowledge based engineering recommend (Knowledge-
Based recommendation) and mixing recommendation (Hybrid recommendation) algorithm.Collaborative filtering is basis
The selection of the hobby before user and the close user of other interest recommends article to user;Content-based recommendation, main
Information retrieval field to be derived from, its principle is the content selecting article by analyzing user in the past, extracts text feature, not
In the article selecting, according to text feature, calculate article similarity, realize recommending;Knowledge based engineering recommend by with user it
Between interaction, such as limit the conditions such as purchasing price interval, brand, function, constantly to approach the in-mind anticipation of user;Mixing is recommended
Algorithm is to be merged three of the above method by way of different, has complementary advantages, and realizes recommending.
In conjunction with the recommendation method of commercial field, research worker proposes multiple commending systems in E-Learning, bag
Include and incorporate body construction in commending system, CB, CF data excavates mixed method etc..In formal middle and primary schools' teaching process
In, during teacher needs to prepare substantial amounts of teaching material and data, to be to give lessons to prepare, and middle and primary schools impart knowledge to students at home, exist
Student with book geographic difference, teachers' level level uneven the problems such as.A lot of commending systems recommend precision low, directly at present
Using collaborative filtering method, lack the understanding understanding to course content, cause to recommend the out-of-date of resource it is recommended that result with prepared
The low problem of the degree of association of task of course, and system is in initial launch, the addition of new user or new resources, lacks history row
For record, it is sparse therefore to easily cause user's history behavioral data, and the problems such as cold start-up;Body is by specific area vocabulary
The basic terminology of table and its relation, and to define, with reference to these terms and relation, the Formal Representation that vocabulary off-balancesheet prolongs rule,
The expert typically requiring this field, to be built and to safeguard, adopts body construction in commending system, needs very big manpower
Material resources put into, accordingly, it would be desirable to one kind can be based on course content, can save labour turnover it is recommended that correlation teaching resource individual character
Change commending system.
Content of the invention
(1) technical problem to be solved
The technical problem that the present invention to be solved is in formal teaching, how fully to excavate course content, analytic instruction provides
Incidence relation between source contents and resource, and how course, teaching resource and teacher's behaviors to combine, realize based on mark
Sign collection and the network teaching resource of teacher's behaviors is recommended.
(2) technical scheme
The invention provides a sharp network teaching resource commending system is it is characterised in that include:
Data builds module, and it builds teacher's behaviors data, tutor model data, course prototype data, resource model number
According to;
Off-line data processing module, it is used for initializing and adjust course prototype data and resource model data, and utilizes
Teacher's behaviors inferred from input data teacher's identity, according to the degree of association between teacher's behaviors data computing resource, according to resource model number
According to the similarity between computing resource, associating between course and resource is calculated according to resource model data and course prototype data
Degree;
Online recommending module, it is associated with resource using the similarity between the degree of association between resource, resource, course
Degree, the Dynamic profiling of teacher recommend resource online, always according to teacher to the feedback recommendation resource tag recommending resource, and by teacher
The behavioral data producing is sent to data and builds module, to update teacher's behaviors data.
Present invention also offers a kind of network teaching resource recommends method it is characterised in that including:
Data construction step, it builds teacher's behaviors data, tutor model data, course prototype data, resource model number
According to;
Off-line data process step, it is used for initializing and adjust course prototype data and resource model data, and utilizes
Teacher's behaviors inferred from input data teacher's identity, according to the degree of association between teacher's behaviors data computing resource, according to resource model number
According to the similarity between computing resource, associating between course and resource is calculated according to resource model data and course prototype data
Degree;
Online recommendation step, it is associated with resource using the similarity between the degree of association between resource, resource, course
Degree, the Dynamic profiling of teacher recommend resource online, always according to teacher to the feedback recommendation resource tag recommending resource, and according to religion
The behavioral data that teacher produces updates teacher's behaviors data.
(3) beneficial effect
From technique scheme as can be seen that the invention has the advantages that:
Characteristic for teacher users and demand, screening from vast resources storehouse is suitable to the teaching resource of teacher, overcomes net
The problem of network resources bank information overload.
1st, the present invention adopts Knowledge Decision-making tree construction, is using the formatting structure of books according to teaching, is designed, with
When, using the paths in decision tree, to represent the Dynamic profiling of tutor model, to screen network of relation under this tree
Resource, can improve precision and the quality of recommendation.
2nd, the present invention is using recommendation method based on content and label, can be prevented effectively from that cold start-up data is sparse to ask
Topic, the impact bringing to commending system.
3rd, the present invention uses association rule mining method, is that operation at short notice has purpose even based on same teacher
Continuous property it is assumed that be translated into can using association rule mining method process problem, calculate teaching resource between pass
Connection degree, in the case of reducing Sparse, finds the difficult impact of neighbour using collaborative filtering.
4th, the present invention uses the method such as latent semantic analysis, statistics, cluster, by teacher's behaviors record, to course label
Carry out filtering extension with resource tag, with the increase of behavior record, the course label in system and resource tag can become to get over
Come more stable and accurate, effectively reduce based on content recommendation method in the inaccuracy extracting key word.
Brief description
Fig. 1 is the structured flowchart of network teaching resource commending system in the present invention;
Fig. 2 is the data structure schematic diagram of tutor model module in the present invention;
Fig. 3 (a) is the data structure schematic diagram of course prototype module in the present invention;
Fig. 3 (b) is the explanation schematic diagram of Fig. 3 (a) symbol;
Fig. 4 is the data structure schematic diagram of resource model module in the present invention;
Fig. 5 (a) is the workflow schematic diagram of webcrawler module in the present invention;
Fig. 5 (b) is the workflow schematic diagram of Chinese version analysis module of the present invention;
Fig. 6 is the workflow schematic diagram that in the present invention, resource mixes recommending module;
Fig. 7 be network teaching resource in the present invention recommendation method in processed offline part method flow diagram;
Fig. 8 (a) shows the method flow diagram of online treatment part in the recommendation method of inventive network teaching resource;
Fig. 8 (b) shows the detailed process schematic diagram of the partly middle course selection of online treatment in the present invention.
Specific embodiment
For making the object, technical solutions and advantages of the present invention become more apparent, below in conjunction with specific embodiment, and reference
Accompanying drawing, the present invention is described in further detail.
As shown in figure 1, the invention discloses a kind of commending system of network teaching resource, this system is divided into data to build mould
Block, off-line data processing module and online recommending module, by teacher's behaviors logging modle 101, tutor model 102, course prototype
103rd, resource model 104, teachers'action describe reasoning module 105, resource associations degree computing module 110, course label adjustment mould
Block 106 and 107, resource tag adjusting module 108 and 109, course resources calculation of relationship degree module 111, resource Similarity measures
Module 112, resource mixing recommending module 113, label recommendations module 114 and UI interactive module 115, totally 15 module compositions.Its
In, teacher's behaviors logging modle 101, tutor model 102, course prototype 103 and resource model 104 constitute data and build module,
It is the data basis of system;Teachers'action describes reasoning module 105, course label adjusting module 106 and 107, resource tag and adjusts
Mould preparation block 108 and 109, resource associations degree computing module 110, course resources calculation of relationship degree module 111, resource Similarity measures
Module 112 constitutes off-line data processing module, and off-line data processing module builds module from data and obtains corresponding data, carries out
Analytical calculation;Resource mixing recommending module 113, label recommendations module 114 and UI interactive module 115 constitute online recommending module,
Online recommending module on the basis of off-line module analysis, for different behaviors, searching resource and money in UI interface for the teacher
Source label, and these behaviors are collected in the teacher's behaviors database of record that data builds module.Separately below to these moulds
Block is introduced.
Teacher's behaviors record 101, for recording teacher in system platform, the various operations carrying out, are personalized recommendation
Data basis are provided.According to the different behaviors of teacher, teacher's behaviors record can be divided into teacher resource behavior record, teachers ' teaching
Task record and teacher's tag behavior record.Teacher resource behavior record is to record teacher in UI interactive module 115, to resource
Operation behavior, including downloading, scoring, collect, browse, as shown in table 1, each teacher has a teacher ID, each resource
Also there is resource ID, in teacher resource behavior record, associated storage teacher is to when the behavior of resource, teacher ID, resource ID and operation
Between;Teachers ' teaching task record refers to teacher in UI interactive module 115, the Instructional Design carrying out and preparing lessons operation, as table 2 institute
Show, wherein associated storage teacher ID, course ID and teaching task content and operating time;Teacher's tag behavior record refers to UI
In interactive module 115, the operation that teacher is carried out to the resource tag recommended, including selection, modification, delete, as shown in table 3, its
Middle associated storage teacher ID, resource ID, recommend label, operation after label and operating time.
Table 1 teacher resource behavior record
Table 2 teachers ' teaching task record
Table 3 teacher's tag behavior record
Fig. 2 shows the data structure schematic diagram of tutor model module 102 in the present invention.As shown in Fig. 2 its data structure
Including static description 201 and Dynamic profiling 202.Static description 201 refers to the essential information of user, and including teacher ID, (system is divided
Join), name, sex, address, education degree and Email etc.;Dynamic profiling 202 refers to the attribute of the taught journey of teacher, demarcates
Its position in course prototype knowledge tree, attribute includes grade, subject, version, unit and course, and asking of attribute is father and son pass
System, the knowledge tree of course prototype 103 is described in detail to this.Static description 201 is when teacher registers, and needs the letter completing
Breath.The acquisition modes of Dynamic profiling 202 include two kinds, and one is the display information feedback after teacher logs in platform, carries out online
Teaching task designs, and this process needs the teacher logging in select the attribute of course;Two be system according to teacher resource behavior record,
To infer the Dynamic profiling of teacher, this part describes in reasoning 105 in teachers'action and is discussed in detail.
Fig. 3 (a) shows the data structure schematic diagram of course prototype 103 in the present invention;Fig. 3 (b) shows in the present invention
The explanation schematic diagram of symbol in course prototype.As shown in Fig. 3 (a) and 3 (b), described course prototype 103 is included by knowledge tree 301
With course label 302.Knowledge tree is decision tree structure, according to curriculum attribute in teaching books, including grade, subject, version, list
Unit and course, domestic primary and secondary teaching material layout is become tree, the 0th layer of corresponding root node, for judging grade belonging to course, the
The judgement of 1 layer of corresponding affiliated subject of course, the judgement of the 2nd layer of corresponding affiliated version of course, the 3rd layer of corresponding course said units
Judge, the judgement of the 4th layer of corresponding place course, the 5th layer is leaf node, for each course specific name of books of imparting knowledge to students, each layer
Word between line, is decision condition.Decision tree structure should be readily appreciated that and realizes, can be to large-scale number within the relatively short time
Make feasible and respond well result according to source, the Dynamic profiling of tutor model can be positioned using hierarchical structure, and make to push away
It is more accurate to recommend.Course label, for representing the content characteristic of course, its data structure is word dynamic set, such as seven grades
People's religion version is geographical《The first segment earth and tellurion》Course label:Latitude, and warp, graticules, longitude, Beijing, map, away from
From, tellurion, the earth, divide, geographical, parallel, initialization is completed by course label adjusting module 106,107, afterwards, is
The teaching task content of system processed offline teacher, continuous adjusting and optimizing course label.Knowledge tree 301 is based on middle and primary schools' teaching book
Decision tree constructed by nationality, root node judges for grade;Grade is from first grade of primary school to high by three, totally ten two grade's classification;Learn
Section includes Chinese language, mathematics, English, politics, biology, history etc., and the foundation of subject decision condition is opened up according to grade
Set by section;Version includes the teaching material version that domestic education of middle and primary schools is used, by the above node, you can decision-making goes out teacher institute
Version taught by the books using, such as Third school grade Chinese language people;Unit condition is the division carrying out according to the catalogue of books, and such as the first volume first is single
Unit, volume two the 3rd unit;After course node judges, you can obtain a certain concrete course.Certainly, during knowledge tree is also not necessarily limited to
Primary school teaching books, it can also be the decision tree that other teaching datas build.303,304,305,306 and 307 in Fig. 3 (b)
The explanation of symbol in corresponding Fig. 3 (a) decision tree.Organize teaching books, such as Third school grade Chinese language people religion according to this decision tree structure
Version first volume second unit《Bullfinch》, you can it is shown in knowledge tree.Course label 302, represents the core content of course, it is initial
Change, completed by the web crawlers 106 in course label adjusting module in Fig. 1 and text analyzing 107, but because text analyzing is logical
Cross computerized algorithm analysis and obtain course label, lack the semantic understanding for content of text, deposit in the course label extracting
In degree of being associated with than relatively low label, i.e. course label noise, need constantly by carrying out to teachers ' teaching task definition point
Analysis, crosses noise filtering, extends course label simultaneously.
Fig. 4 shows the data structure schematic diagram of resource model 104 in the present invention.As shown in figure 4, inclusion resource ID 401,
Title 402, key word 403, description 404 and resource tag 405.Resource ID 401 is default, title 402, key word 403
It is the associated description to resource for the resource owner with description 404;Resource tag 405 represents the content characteristic of resource, and its data is tied
Structure is word dynamic set, such as resource《Earth movements》Resource tag be:White turn, and the earth, revolution, motion }, its initialization,
Completed by the resource content analysis 108 in resource tag adjusting module in Fig. 1 and resource tag analysis 109.With course prototype phase
Same, there is also noise in resource tag, exist in resource tag with resource associations degree than relatively low label, by marking to teacher
Sign behavior record and carry out Algorithm Analysis, cross noise filtering, extended resources label.
Teachers'action describes reasoning module 105, and it uses the Dynamic profiling of two methods reasoning teacher, described two methods
Including:One is the resource content inference method of knowledge based tree:By calculating the resource being related in teacher resource behavior record
The degree of association between label course label corresponding with knowledge leaf nodes, carrys out the Dynamic profiling of reasoning teacher.As:Teacher 1801
Registered address be Hebei province Baoding, the down loading network resource behavior within a period of time as shown in table 1, then computing resource 1,
2nd, the degree of association between 3 resource tag course label corresponding with knowledge leaf nodes, deducibility teacher 1801 is taught, and journey is
High school physicses《Newton's second law》, according to the registered address information of teacher 1801, the version of the taught journey of deducibility teacher 1801
This teaches version for people, and finally can orient the current information of giving lessons of teacher 1801 in knowledge tree is:High school physicses people teaches version the 4th
Chapter《3 Newton's second laws》.Two is Cooperative Reasoning method:Find the neighbour of any active ues, in k neighbour teacher, if certain
Teacher has clear and definite Dynamic profiling, then infer that any active ues have identical Dynamic profiling, such as:Teacher 1802 has and clearly dynamically retouches
State, and have common behavior record to be related to resource with teacher 1801, therefore, can reasoning teacher 1 801 and teacher 1 802 dynamic
Description is identical.By the reasoning of Dynamic profiling, you can recommend this section or the related resource of next section, it is to avoid course before recommending
Related resource so that recommend more accurate.
Course label adjusting module, input course prototype and teachers ' teaching task record, are divided by web crawlers and text
Analysis, exports knowledge tree leaf node corresponding course label.Course label adjusting module, including webcrawler module 106 and literary composition
This analysis module 107, when described webcrawler module 106 is used for the initialization of course label, crawls in the webpage that course is associated
Hold, described text analysis model 107 is used for extracting the content characteristic of input text, and described input text includes webcrawler module
Teachers ' teaching task record in web page contents and teacher's behaviors logging modle 101 that 106 courses crawling are associated.Specifically
Ground, described text analysis model 107 according to processing procedure, dynamic adjustment course label, if initialization procedure, then extract webpage
Content characteristic, merges Web Page Key Words by course, as course label, if makeover process, then extracts in teachers ' teaching task
Hold feature, using clustering method, the label liveness method of falling row and latent semantic analysis, course label is filtered and extend.Institute
State the quantity that the course of certain label in clustering method analysis, will appear from number of times and delete from course label more than the label of certain threshold value
Remove, this is because the label more than occurrence number, there is no good separating capacity, such as occur in the course label of a lot of courses
" teaching ", then this word can not be used for distinguishing these courses it is therefore desirable to delete.The label liveness extraction of falling discharge method teacher teaches
Learn the label of task definition, and count the frequency that each label occurs, as label liveness, by label low for liveness, from class
Delete in journey label.Latent semantic analysis are used for Give lecture-label matrix, using matrix disassembling method, the spy of solution matrix
Levy vector, analyze characteristic vector using distance metric method, solve the neighbour of label, to extend course tag set.
Fig. 5 (a) and Fig. 5 (b) respectively illustrates described webcrawler module 106 and the workflow of text analysis model 107
Journey schematic diagram.As shown in Fig. 5 (a), the workflow of webcrawler module is specific as follows shown:
Step 501:Select teaching books, such as one grade Chinese language people teaches version;
Step 502:Extract the title of each course in teaching books;
Step 503:Based on search engine technique, using course name as key word, search for the related web page of course;
Step 504:Obtain web page listings and chained address;
Step 505:Crawl the web page contents in list link;
Step 506:Judge whether all of teaching books terminate, if so, then enter next stage, otherwise, jump to step
Rapid 501, select next books of imparting knowledge to students.
As shown in Fig. 5 (b), the workflow of text analysis model is specific as follows shown:
Step 507:Extracted from the related web page contents of course and teachers ' teaching task record using TF-IDF algorithm and close
Keyword, if frequency TF that certain word or phrase occur in an article is high, and the reverse frequency occurring in other articles
Rate IDF is low then it is assumed that this word or phrase have good class discrimination ability, and therefore extracting it is key word;
Step 508:Merge key word by course, will extract from different web pages and/or teachers ' teaching task record
A set merged in the key word of same course, is such as obtained the key word of two webpages of course 1, (course 1, net by step 1
Page 1):The earth, and revolution, white turn, the sun, multicolored vaginal discharge, round the clock }, (course 1, webpage 2):{ revolution, the sun, tellurion, the earth, ground
Reason, multicolored vaginal discharge, round the clock }, merge into, course 1:The earth, and revolution, the sun, white turn, tellurion, geographical, multicolored vaginal discharge, round the clock };
Step 509:According to input content of text source, adjust course label, if the web page contents that web crawlers crawls,
Then directly key word is exported course label model 103 as course label;If the key word that teachers ' teaching task obtains
Set, then filtered to the course label of course prototype and extended.Filter and the method for extension has label liveness to fall to arrange, dive
In semantic analysis, clustering method etc..Label liveness falls to arrange, and refers to the keyword set obtaining according to teachers ' teaching task, system
The number of times that meter wherein each key word occurs, filters the label not occurring or infrequently occurring in course label;Potential applications divide
Analysis, the method using singular value decomposition is decomposed to course-label matrix, calculates the neighbour of characteristic vector, to find label
Between similarity, and not word in course label high with course label similarity is added in course label, realizes
Expanded function;Clustering method, then be statistics everyday words, for example, all there is " teaching ", it is unable to table in multiple course tag sets
Show the content characteristic of course, therefore filter this out.
Resource tag adjusting module, the tag behavior record of teacher and resource model as input, using clustering method, mark
Sign the liveness method of falling row and potential applications method, resource tag is filtered and extends.Resource tag adjusting module includes resource
Content analysis module 108 and resource tag analysis module 109.Resource content analysis module 108 is by DF Algorithm Analysis resource
Title, key word, description, and therefrom filter out resource tag, specifically, the document frequency DF if there is certain word is high,
Then this word does not have good separating capacity, then removed, using remaining word as resource tag;Resource tag analyzes mould
Block 109 is used for the resource tag in the resource tag and teacher's tag behavior record that resource content analysis module 108 is filtered out
After being optimized adjustment, output this to resource model.If in resource tag initialization procedure, being directly output to resource
In model, table 4 is the resource model after resource tag initialization;If calculating teacher's tag behavior record gained, then using mark
Sign liveness row, latent semantic analysis and clustering method statistics, filter and extended resources label.Label liveness falls to arrange, system
The number of times that in meter resource tag, each key word occurs, the label not occurring in filtered resources label or infrequently occurring;Potential
Semantic analysis, the method resource-label matrix using singular value decomposition is decomposed, and calculates the neighbour of characteristic vector, to find
Similarity between resource tag, and by and not word in resource tag high with resource tag similarity, it is added to resource
In label;Clustering method, then be statistics everyday words, if it can not represent the content characteristic of resource, be deleted.
Table 4 resource model and the resource tag of extraction
Resource associations degree computing module 110, inputs teacher resource behavior record, the degree of association between output resource, method is to recognize
For same teacher behavior at short notice, there is seriality, that is, same teacher has the resource of behavior record at short notice is to close
Connection, problem is converted into the degree of association between computing resource, is solved using association rule mining method.This module base
Have successional it is assumed that analyzing the degree of association between resource in same teacher behavior operation at short notice:I.e. same religion
Teacher at short notice, can carry out browsing, download, score or collecting to some resources, then these resources are related, should
Teacher during this period of time, has the resource of behavior record, is converted into the note once having transaction number in system platform for the teacher
Record, all teacher resource behavior records are all converted into transaction record, in this way such that it is able to obtain teacher in system platform
Resource transaction inventory, be converted into association rule mining problem, two resources can be calculated using Apriori algorithm and occur simultaneously
Behavior record bar number, therefore, according to the degree of association between formula 1 computing resource.
(formula 1)
Num (res1 ∩ res2) represents the number of times occurring during resl and res2 department;
Num (resl ∪ res2) represents the total degree that resl or res2 occurs.
Assume that the teacher resource behavior record that teacher 1801 and teacher 1802 are carried out to resource 1,2,3 is as shown in table 1.Teacher
1801 at short notice, has download behavior to resource 1,2,3, and being converted into transaction record is:Transaction number is 1 the Resources list
{ 1,2,3 };Teacher 1802 at short notice, has download behavior, but the behavior to resource 1 is at second day to resource 2,3
Operation, being converted into transaction record is:Transaction number is 2 the Resources list { 2,3 }, and transaction number is 3 the Resources list { 1 }.As
It is 20, i.e. Num (res2 ∩ res3)=20 that really multiple users have such behavior record bar number at short notice to resource 2,3,
And behavior record bar number resource 2 is 25, behavior record bar number resource 3 is 30, then
Num (res2 ∪ res3)=Num (res2)+Num (res3)-Num (res2 ∩ res3)=25+30-20=35, resource
2nd, 3 degree of association be relevance=20/35=0.571, if set degree of association threshold value as 0.5≤relevance≤1 then it is assumed that
Resource 2,3 is associated, and the degree of association is 0.571.
Course resources label calculation of relationship degree module 111, course label and resource tag as input, using distance metric
Method calculates the degree of association of course label and resource tag, using k near neighbor method or the setpoint distance threshold filtering degree of association relatively
Low resource, the degree of association between final output course and resource.Distance metric method choice formula 2, calculates course and resource
Between the degree of association, not less than certain threshold value then it is assumed that be association.
(formula 2)
IntS (crs, res) represents the element of set { tag | tag ∈ course label crs, and tag ∈ resource tag res }
Number, SupS (crs, res) represents the element of set { tag | tag ∈ course label crs, or tag ∈ resource tag res }
Number.
Resource similarity computing module 112, input resource tag, using between distance metric method computing resource away from
From using the relatively low resource of k near neighbor method or setpoint distance threshold filtering similarity, between final output resource and resource
Similarity.Distance metric method choice formula 3, the similarity of computing resource 1 and resource 2, not less than certain threshold value then it is assumed that being
Similar.Tally set as resource 1 is { latitude, China, multimedia teaching, longitude, map, warp, graticules, parallel }, money
The tally set in source 2 is { latitude, the world, multimedia teaching, longitude, map, warp, graticules, parallel }, then sim=7/9=
0778.
(formula 3)
IntS (crs, res) represents the unit of set { tag | tag ∈ resource tag resl, and tag ∈ resource tag res2 }
Plain number, SupS (crs, res) represents the element of set { tag | tag ∈ resource tag resl, or tag ∈ resource tag res2 }
Number.
Fig. 6 shows the workflow schematic diagram of resource mixing recommending module 113 in the present invention.As shown in fig. 6, online to
User carries out teaching resource recommendation, the data being obtained using processed offline, and according to tutor model Dynamic profiling and course money
The degree of association in source, calculates and resource Similarity measures in conjunction with resource associations degree, and resource is recommended in mixing.
Step 601:Judge whether teacher's behaviors are searching resource, if so, then jump to step 602, otherwise jump to step
Rapid 603;
Step 602:According to the query word of teacher's input, carry out Keywords matching with the resource tag in resource model, obtain
To the Resources list high with Keywords matching degree, when entering certain resource in the Resources list as teacher, such as teacher's search keyword
" li po ", system, in all resource models, carries out Keywords matching, obtains the Resources list:Resource 1 li po .jpg, resource
2 li po .doc, resource 3 will enter wine .doc etc., and when teacher clicks on certain resource, such as resource 3 will enter wine .doc, enter into
During the detailed description interface of this resource, jump to step 604;
Step 603:Process the result obtaining from course resources calculation of relationship degree module 111, filter out and tutor model
Dynamic profiling (grade, subject, version, unit, course) the higher resource of the degree of association, obtains the Resources list;
Step 604:In conjunction with resource associations degree computing module 110 and resource similarity computing module 112, find with resource or
The higher resource of resource in the Resources list, the degree of association and similarity, according to default weights, the degree of association to these resources and
Similarity is weighted, and obtains related resource list, and the weight of each resource, such as with resource 1 degree of association and similarity
Higher resource 2, its degree of association is 0.8, and similarity is 0.5, and default weights are respectively 0.6,0.4, then the weight of resource 2 is
0.8*0.6+0.5*0.4=0.68:
Step 605:Row fall according to weight height it is recommended that teaching resource.
Label recommendations module 114, is after teacher uses resource, recommends the respective labels of this resource to it, obtain teacher
Feedback information, such as teacher downloads《World map .jpg》, then to its recommend " world ", " map ", " longitude ", " latitude ",
" warp ", " parallel " etc., teacher can modify, select to the label recommended, deletion action.
UI interactive module 115 is responsible for interacting between teacher, according to the different behaviors of teacher, realizes personalized recommendation,
Meanwhile, collect teacher's various operation behaviors in systems, these behavior records are exported, is saved in the teacher's behaviors of data Layer
In record.
The invention allows for a kind of recommendation method of network teaching resource.The method includes processed offline and online treatment
Two parts flow process.
Processed offline is to build data and the model of module by off-line analysiss data, recommends to provide mediant for online
According to, and analytical data builds the teacher's behaviors data that module collection arrives, and continues to optimize each model that data builds module.
Fig. 7 shows the method and step flow process of processed offline part in the recommendation method of network teaching resource in the present invention
Figure.As shown in fig. 7, modules process can separate be run in described commending system, suitable Distributed Parallel Computing, can
For processing big data, system is conducive to extend.The step of processed offline part is as follows, it can be seen from figure 7 that each step
Suddenly independence and out-of-order relation are mutually asked:
Step 701:Obtain teacher resource behavior record from teacher's behaviors logging modle 101, pushed away using teachers'action description
Reason module 105 makes inferences to teachers'action description, obtains the Dynamic profiling of teacher;Simultaneously using resource associations degree computing module
Resource associations degree in 110 computing resource models;
Step 702:Obtain course name from course prototype 103, by the web crawlers 106 of course label adjusting module
With text analyzing 107, the course label of described course is initialized;Obtain teacher's religion from teacher's behaviors logging modle 101
Learn task record, by text analysis model 107, course label is filtered and extends;
Step 703:Obtain resource content from resource model 104, including asset title, key word and description, by money
The resource content analysis 108 of source label adjusting module and resource tag analysis 109, initialize to resource tag;From teacher
Behavior record module 101 obtains teacher's tag behavior record, by resource tag analysis module 109, resource tag is carried out
Filter and extension;
Step 704:Course prototype data and resource model number is obtained respectively from course prototype 103 and resource model 104
According to both combine, and by course resources calculation of relationship degree module 111, calculate the course resources degree of association;
Step 705:From resource model 104, obtain resource model data, by resource similarity computing module 112, count
Calculate the similarity asked of resource.
Fig. 8 (a) shows the method flow diagram of online treatment part in the present invention.Online treatment part is in UI circle,
By with teacher ask interact, realize resource recommendation and label recommendations, as shown in figure 8, step is as follows:
Step 801:Judge whether user registers:If so, then log on as teacher;Otherwise, register user, system registry is newly used
The tutor model at family, that is, register the essential information of new user, including teacher ID (system distribution), name, sex, address, education
Degree and Email etc.;
Step 802:Judge the behavior of teacher:After teacher's logging in network platform, if preparing teaching task, such as online teaching
Design, prepare lessons online, then platform carries out step 803 course selection, if searching resource, jumps to step 804, if carrying out it
His behavior, such as manages personal resource, modification information etc., then jumps to step 806;
Step 803:Select course association attributes, and dynamically retouching of teacher is explicitly obtained according to the selected course of teacher
State, including grade, subject, version, unit and course, change tutor model 102, jump to step 807;
Step 804:Teacher is actively entered search terms, searching resource, and system is according in input key word, with resource model
Resource tag carry out Keywords matching, obtain the Resources list high with Keywords matching degree, such as teacher search keyword " Lee
In vain ", system, in all resource models, carries out Keywords matching, obtains the Resources list:Resource 1 li po .jpg, resource 2
Li po .doc, resource 3 will enter wine .doc etc.;
Step 805:Teacher obtains with reference to resource associations degree computing module 110 to certain resource operation in the Resources list, system
Resource degree of association and the resource similarity that obtains of resource similarity computing module 112, using mixing recommending module 113, find
With the resource in resource or the Resources list, the degree of association and the higher resource of similarity, according to default weights, to these resources
The degree of association and similarity are weighted, and obtain related resource list, and the weight of each resource, such as with resource 1 degree of association
The resource 2 higher with similarity, its degree of association is 0.8, and similarity is 0.5, and default weights are respectively 0.6,0.4,
Then the weight of resource 2 is 0.8*0.6+0.5*0.4=0.68;Recommend teaching resource according to weight height, jump to step
Rapid 808;
Step 806:System extracts the Dynamic profiling of tutor model 102;
Step 807:Teaching resource is recommended by resource recommendation module:Resource recommendation module is dynamic according to tutor model 102
Description, in conjunction with resource associations degree computing module 110, course resources calculation of relationship degree module 111 and resource similarity computing module
112, processed the result obtaining from course resources calculation of relationship degree module 111 using mixing recommending module 113, filter out and teach
Teacher's model Dynamic profiling (grade, subject, version, unit, course) the higher resource of the degree of association, obtains the Resources list;In conjunction with money
Source calculation of relationship degree module 110 and resource similarity computing module 112, find and the resource in resource or the Resources list, association
Degree and the higher resource of similarity, according to default weights, are weighted to the degree of association and similarity of these resources, obtain
To related resource list, and the weight of each resource;Recommend teaching resource according to weight height, jump to step 808;
Step 808:System obtains teacher resource behavior feedback information, such as downloads, browses, scores, and the money by teacher
Source behavior record and teaching task record are saved in teacher's behaviors logging modle 101;
Step 809:Fed back according to teacher resource behavior, resource tag recommending module recommends related resource tag, such as teaches
Teacher has downloaded resource《Heliocentric theory .jpg》, then by resource tag { meaning, Copernius, theoretical, core content, the day heart of this resource
Say, open, increase, the visual field recommend teacher;
Step 810:Collect the feedback to label for the teacher, include selecting, the behavior such as delete, change, and by teacher's row of labels
It is saved in teacher's behaviors logging modle 101 for record, such as teacher is to the resource recommended《Heliocentric theory .jpg》Resource tag, select
" heliocentric theory " and " Copernius ", deletes " meaning ", " theoretical ", " open ", " increasing ", " visual field ";
Step 811:Judge whether teacher exits, if so, then exit;Otherwise, jump procedure 802.
Output to data is all built in the teacher's behaviors record of module, supplies by the operation behavior that teacher is carried out in UI module
Processed offline.
Fig. 8 (b) shows the course selection detailed process of step 803 in the present invention.As shown in Fig. 8 (b), this step bag
Include:
Step 812:Select grade;
Step 813:Select subject;
Step 814:Select version;
Step 815:Select unit;
Step 816:Select specific course.
System and method disclosed by the invention can in magnanimity teaching resource fast and accurately by resources of interest on demand
Recommend teacher users, effectively overcome network resources system problem of information overload, reduce the workload of teacher, set for completing teaching
The teaching task such as counting, prepare lessons provides facility.Meanwhile, the present invention can effectively reduce the shadow that cold start-up data Sparse Problems bring
Ring, improve the interactivity that system is asked with teacher, be that the personalized recommendation of network teaching resource provides a kind of pervasive method.
Particular embodiments described above, has carried out detailed further to the purpose of the present invention, technical scheme and beneficial effect
Describing in detail bright it should be understood that the foregoing is only the specific embodiment of the present invention, being not limited to the present invention, all
Within the spirit and principles in the present invention, any modification, equivalent substitution and improvement done etc., should be included in the protection of the present invention
Within the scope of.
Claims (6)
1. a kind of network teaching resource commending system is it is characterised in that include:
Data builds module, and it builds teacher's behaviors data, tutor model data, course prototype data, resource model data;
Off-line data processing module, it is used for initializing and adjust course prototype data and resource model data, and utilizes teacher
Behavioral data infers teacher's identity, according to the degree of association between teacher's behaviors data computing resource, according to resource model data meter
Calculate the similarity between resource, calculate the degree of association between course and resource according to resource model data and course prototype data;
Online recommending module, its utilize the degree of association of similarity, course and resource between the degree of association between resource, resource,
The Dynamic profiling of teacher recommends resource online, always according to teacher to the feedback recommendation resource tag recommending resource, and teacher is produced
Raw behavioral data is sent to data and builds module, to update teacher's behaviors data;
Described teacher's behaviors data includes:Teacher resource behavior record, teachers ' teaching task record and teacher's tag behavior record,
Wherein teacher resource behavior record includes the behavior to resource for the teacher, and teachers ' teaching task record includes the teaching that teacher carries out and sets
Meter and operation of preparing lessons, teacher's tag behavior record includes the operation to resource tag for the teacher;
Described tutor model data includes static description and the Dynamic profiling of teacher, the essential information of static description teacher of the giving advice, and moves
State description includes the attribute of the taught journey of teacher;
Described course prototype data includes knowledge tree and course label;Described knowledge tree adopts decision tree structure, is according to teaching
The tree that in books, curriculum attribute is lined up;
Described resource model includes Resource Properties and resource tag.
2. the system as claimed in claim 1 is it is characterised in that described off-line data processing module includes:
Teachers'action describes reasoning module, using the resource content inference method of Cooperative Reasoning method and knowledge based tree, according to k
The Dynamic profiling of neighbour teacher, teacher to the resource tag being related in the operation behavior record data of resource and course label it
Between the degree of association, the Dynamic profiling of reasoning teacher;
Resource associations degree computing module, according between same teacher in the given time the operation behavior computing resource to resource
The degree of association;
Course label adjusting module, according to course name in course prototype data from web page crawl course related web page, and therefrom
Extract course key word and initialize described course prototype data as course label;Taught according to teacher in teachers ' teaching task record
Learn task definition feature, using clustering method, the label liveness method of falling row and latent semantic analysis, course label is filtered
And extension;
Resource tag adjusting module, extracts resource tag according to resource content in resource model data, to initialize described resource
Model data;To the resource tag obtaining from teacher's tag behavior record utilize label liveness fall row, latent semantic analysis and
Clustering method counts, and filters and extended resources label;
Course resources calculation of relationship degree module, calculates the degree of association of course label and resource tag using distance metric method, and
Using the relatively low resource of k near neighbor method or the setpoint distance threshold filtering degree of association, the pass between final output resource and course
Connection degree;
Resource similarity computing module, using the distance between distance metric method computing resource label, using k near neighbor method or
The relatively low resource of person's setpoint distance threshold filtering similarity, the similarity between final output resource and resource.
3. the system as claimed in claim 1, wherein, online recommending module includes:
Resource mixes recommending module, according to the query word of teacher's input, obtains the Resources list by mating resource tag, or root
According in tutor model data teachers'action description, obtain with this teachers'action describe the higher resource of the degree of association obtain resource row
Table, is higher than then predetermined association angle value and Similarity value according to the Resource Calculation degree of being associated with the Resources list and similarity
Association and similar resource, compute weighted to the association being obtained and similar resource according to default weights, obtain related money
Source list, weighted value in related resource list is higher than the resource recommendation of predefined weight value to teacher;
Label recommendations module, is that it feeds back the resource tag of this recommendation resource to the operation behavior of recommended resource according to teacher;
Teacher's behaviors logging modle, the behavioral data for producing teacher is sent to data and builds module, with more beginning teacher's row
For data.
4. a kind of network teaching resource recommends method it is characterised in that including:
Data construction step, it builds teacher's behaviors data, tutor model data, course prototype data, resource model data;
Off-line data process step, it is used for initializing and adjust course prototype data and resource model data, and utilizes teacher
Behavioral data infers teacher's identity, according to the degree of association between teacher's behaviors data computing resource, according to resource model data meter
Calculate the similarity between resource, calculate the degree of association between course and resource according to resource model data and course prototype data;
Online recommendation step, its utilize the degree of association of similarity, course and resource between the degree of association between resource, resource,
The Dynamic profiling of teacher recommends resource online, always according to teacher to the feedback recommendation resource tag recommending resource, and according to teacher
The behavioral data producing updates teacher's behaviors data;
Described teacher's behaviors data includes:Teacher resource behavior record, teachers ' teaching task record and teacher's tag behavior record,
Wherein teacher resource behavior record includes the behavior to resource for the teacher, and teachers ' teaching task record includes the teaching that teacher carries out and sets
Meter and operation of preparing lessons, teacher's tag behavior record includes the operation to resource tag for the teacher;
Described tutor model data includes static description and the Dynamic profiling of teacher, the essential information of static description teacher of the giving advice, and moves
State description includes the attribute of the taught journey of teacher;
Described course prototype data includes knowledge tree and course label;Described knowledge tree adopts decision tree structure, is according to teaching
The tree that in books, curriculum attribute is lined up;
Described resource model includes Resource Properties and resource tag.
5. method as claimed in claim 4, wherein, off-line data process step includes:
Using the resource content inference method of Cooperative Reasoning method and knowledge based tree, according to Dynamic profiling, the religion of k neighbour teacher
The degree of association between the resource tag being related in the behavior record of qualified teachers source and course label, the Dynamic profiling of reasoning teacher;And
According to the degree of association between same teacher in the given time the operation behavior computing resource to resource;
According to course name in course prototype data from web page crawl course related web page, and therefrom extract course key word conduct
Course label initializes described course prototype data;According to teachers ' teaching task definition feature in teachers ' teaching task record, make
With clustering method, the label liveness method of falling row and latent semantic analysis, course label is filtered and extended;
Resource tag is extracted according to resource content in resource model data, to initialize described resource model data;To from teacher
The resource tag that tag behavior record obtains utilizes label liveness to fall to arrange, latent semantic analysis and clustering method count, and filters
With extended resources label;
Calculate the degree of association of course label and resource tag using distance metric method, and using k near neighbor method or set away from
From the relatively low resource of the threshold filtering degree of association, the degree of association between final output resource and course;
Using the distance between distance metric method computing resource label, using k near neighbor method or setpoint distance threshold filtering
The relatively low resource of similarity, the similarity between final output resource and resource.
6. method as claimed in claim 4, wherein, online recommendation step includes:
According to the query word of teacher's input, obtain the Resources list by mating resource tag, or according in tutor model data
Teachers'action description, obtain describing the higher resource of the degree of association with this teachers'action and obtain the Resources list, then according to resource
Resource Calculation degree of being associated with list and similarity are higher than association and the similar resource of predetermined association angle value and Similarity value,
According to default weights, the association being obtained and similar resource are computed weighted, obtain related resource list, by related money
In the list of source, weighted value is higher than the resource recommendation of predefined weight value to teacher;
According to teacher, the operation behavior of recommended resource is fed back with the resource tag of this recommendation resource;
Teacher's behaviors data is updated according to the behavioral data that teacher produces.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410093793.6A CN103886054B (en) | 2014-03-13 | 2014-03-13 | Personalization recommendation system and method of network teaching resources |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410093793.6A CN103886054B (en) | 2014-03-13 | 2014-03-13 | Personalization recommendation system and method of network teaching resources |
Publications (2)
Publication Number | Publication Date |
---|---|
CN103886054A CN103886054A (en) | 2014-06-25 |
CN103886054B true CN103886054B (en) | 2017-02-15 |
Family
ID=50954946
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201410093793.6A Expired - Fee Related CN103886054B (en) | 2014-03-13 | 2014-03-13 | Personalization recommendation system and method of network teaching resources |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN103886054B (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2024033951A1 (en) * | 2022-08-11 | 2024-02-15 | Social Things S.R.L. | Computer- implemented method for providing a recommendation to a teacher user for creating a personalized teaching course |
Families Citing this family (35)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104504517A (en) * | 2014-12-17 | 2015-04-08 | 天脉聚源(北京)教育科技有限公司 | Teaching data management device for intelligent teaching system and management method thereof |
CN104504948A (en) * | 2014-12-17 | 2015-04-08 | 天脉聚源(北京)教育科技有限公司 | Method and device for displaying pushed course for intelligent teaching system |
CN104680453B (en) * | 2015-02-28 | 2017-10-03 | 北京大学 | Course based on student's attribute recommends method and system |
CN104952009A (en) * | 2015-04-23 | 2015-09-30 | 阔地教育科技有限公司 | Resource management method, system and server and interactive teaching terminal |
CN105183782A (en) * | 2015-08-14 | 2015-12-23 | 广东能龙教育股份有限公司 | Analysis recommendation method based on APP platform and data mining |
CN105117460A (en) * | 2015-08-21 | 2015-12-02 | 湖南亿谷科技发展股份有限公司 | Learning resource recommendation method and system |
CN105512256A (en) * | 2015-12-01 | 2016-04-20 | 深圳拼课邦科技有限公司 | Method and device for pushing lecturer information |
CN105719206A (en) * | 2016-01-01 | 2016-06-29 | 河南牧业经济学院 | Higher-education teaching resource configuration and analysis system |
CN105894091A (en) * | 2016-03-31 | 2016-08-24 | 湘潭大学 | Test question difficulty factor knowledge discovery method based on collaborative decision-making mechanism |
CN106528656B (en) * | 2016-10-20 | 2019-08-20 | 杭州新百锐基业科技股份有限公司 | A kind of method and system for realizing that course is recommended based on student's history and real-time learning state parameter |
CN107122399A (en) * | 2017-03-16 | 2017-09-01 | 中国科学院自动化研究所 | Combined recommendation system based on Public Culture knowledge mapping platform |
CN107133307A (en) * | 2017-04-28 | 2017-09-05 | 太仓诚泽网络科技有限公司 | A kind of multimedia teaching video playback interface supplying system |
CN107395725B (en) * | 2017-07-26 | 2020-08-04 | 广东小天才科技有限公司 | Application program recommendation method, application program recommendation device and electronic equipment |
CN109427218B (en) * | 2017-08-25 | 2021-07-30 | 北京三好互动教育科技有限公司 | Online education system and method |
CN107908650B (en) * | 2017-10-12 | 2019-11-05 | 浙江大学 | Knowledge train of thought method for auto constructing based on mass digital books |
CN107731034A (en) * | 2017-11-09 | 2018-02-23 | 北京市农林科学院 | A kind of remote education terminal, service end and Distance Education Resources recommend method |
CN109801525B (en) * | 2017-11-17 | 2021-05-14 | 深圳市鹰硕技术有限公司 | Teacher-student multidimensional matching method and system for network teaching |
CN108121785A (en) * | 2017-12-15 | 2018-06-05 | 华中师范大学 | A kind of analysis method based on education big data |
CN108596804A (en) * | 2018-04-28 | 2018-09-28 | 重庆玮宜电子科技有限公司 | Multithreading online education evaluation method |
CN108763342A (en) * | 2018-05-14 | 2018-11-06 | 北京比特智学科技有限公司 | Education resource distribution method and device |
CN109299368B (en) * | 2018-09-29 | 2020-11-24 | 北京思路创新科技有限公司 | Method and system for intelligent and personalized recommendation of environmental information resources AI |
CN109508429B (en) * | 2019-01-30 | 2021-06-15 | 四川省电子信息产业技术研究院有限公司 | Individualized self-adaptive learning recommendation method based on big data analysis of education platform |
CN110060538A (en) * | 2019-04-08 | 2019-07-26 | 上海云之驾科技股份有限公司 | Personalized artificial based on historical data modeling intelligently drives training and practices system and method |
CN110457283A (en) * | 2019-05-24 | 2019-11-15 | 四川信息职业技术学院(广元无线电技工学校) | A kind of teaching comment based on data mining and recommender system and its method |
CN112148964B (en) * | 2019-06-29 | 2022-11-18 | 阿里巴巴集团控股有限公司 | Information processing and recommending method, system and equipment |
CN110457493A (en) * | 2019-07-24 | 2019-11-15 | 安徽新知数媒信息科技有限公司 | A kind of interactive Rich Media's intelligent recommendation and the method locally called |
CN111292212A (en) * | 2020-03-04 | 2020-06-16 | 湖北文理学院 | Personalized thinking political affairs education system |
CN112765374A (en) * | 2020-07-27 | 2021-05-07 | 上海斐杰教育科技有限公司 | Education resource screening system and method for information push |
CN112069423A (en) * | 2020-08-14 | 2020-12-11 | 京华信息科技股份有限公司 | Information recommendation method and device, storage medium and computer equipment |
CN112182413B (en) * | 2020-12-01 | 2021-02-19 | 三盟科技股份有限公司 | Intelligent recommendation method and server based on big teaching data |
CN112287239B (en) * | 2020-12-30 | 2021-03-19 | 平安科技(深圳)有限公司 | Course recommendation method and device, electronic equipment and storage medium |
CN114595932A (en) * | 2021-04-22 | 2022-06-07 | 朱洪东 | Service resource allocation method based on big data and readable storage medium |
CN116431895A (en) * | 2023-02-15 | 2023-07-14 | 武汉博晟安全技术股份有限公司 | Personalized recommendation method and system for safety production knowledge |
CN117009829A (en) * | 2023-10-07 | 2023-11-07 | 成都华栖云科技有限公司 | Similarity recognition method and device for recorded and broadcast courses |
CN117407594B (en) * | 2023-12-12 | 2024-03-12 | 山东大学 | Book information recommendation system |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101169906A (en) * | 2006-10-24 | 2008-04-30 | 天象网络技术(上海)有限公司 | Integrated teaching procedure |
CN102063459A (en) * | 2010-10-20 | 2011-05-18 | 袁昱明 | Distance educational system supported by digital resource management and integration technology |
CN103116657A (en) * | 2013-03-11 | 2013-05-22 | 中国科学院自动化研究所 | Individual searching method of network teaching resource |
CN103455576A (en) * | 2013-08-22 | 2013-12-18 | 西安交通大学 | Thinking-map-based e-learning resource recommendation method |
CN103544663A (en) * | 2013-06-28 | 2014-01-29 | Tcl集团股份有限公司 | Method and system for recommending network public classes and mobile terminal |
Family Cites Families (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US7885913B2 (en) * | 2007-03-28 | 2011-02-08 | Yahoo! Inc. | Distributed collaborative knowledge generation system wherein students perform queries using a dynamic knowledge database and retrieved subsets of data are shared with multiple users on the web |
US8190590B2 (en) * | 2007-08-15 | 2012-05-29 | Martin Edward Lawlor | System and method for the creation and access of dynamic course content |
US20110010210A1 (en) * | 2009-07-10 | 2011-01-13 | Alcorn Robert L | Educational asset distribution system and method |
-
2014
- 2014-03-13 CN CN201410093793.6A patent/CN103886054B/en not_active Expired - Fee Related
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN101169906A (en) * | 2006-10-24 | 2008-04-30 | 天象网络技术(上海)有限公司 | Integrated teaching procedure |
CN102063459A (en) * | 2010-10-20 | 2011-05-18 | 袁昱明 | Distance educational system supported by digital resource management and integration technology |
CN103116657A (en) * | 2013-03-11 | 2013-05-22 | 中国科学院自动化研究所 | Individual searching method of network teaching resource |
CN103544663A (en) * | 2013-06-28 | 2014-01-29 | Tcl集团股份有限公司 | Method and system for recommending network public classes and mobile terminal |
CN103455576A (en) * | 2013-08-22 | 2013-12-18 | 西安交通大学 | Thinking-map-based e-learning resource recommendation method |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2024033951A1 (en) * | 2022-08-11 | 2024-02-15 | Social Things S.R.L. | Computer- implemented method for providing a recommendation to a teacher user for creating a personalized teaching course |
Also Published As
Publication number | Publication date |
---|---|
CN103886054A (en) | 2014-06-25 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN103886054B (en) | Personalization recommendation system and method of network teaching resources | |
CN108021616B (en) | Community question-answer expert recommendation method based on recurrent neural network | |
CN103064945B (en) | Based on the Situational searching method of body | |
CN106802915A (en) | A kind of academic resources based on user behavior recommend method | |
CN106815297A (en) | A kind of academic resources recommendation service system and method | |
CN105740366A (en) | Inference method and device of MicroBlog user interests | |
CN103425763B (en) | User based on SNS recommends method and device | |
CN106156354A (en) | A kind of education resource commending system | |
CN112966091B (en) | Knowledge map recommendation system fusing entity information and heat | |
CN108509517B (en) | Streaming topic evolution tracking method for real-time news content | |
CN110110225A (en) | Online education recommended models and construction method based on user behavior data analysis | |
CN109063147A (en) | Online course forum content recommendation method and system based on text similarity | |
CN111143672A (en) | Expert specialty scholars recommendation method based on knowledge graph | |
CN111191099B (en) | User activity type identification method based on social media | |
CN115329200A (en) | Teaching resource recommendation method based on knowledge graph and user similarity | |
CN107045533A (en) | Educational resource based on label recommends method and system | |
Yin et al. | Detecting latent topics and trends in blended learning using LDA topic modeling | |
Zhong et al. | Design of a personalized recommendation system for learning resources based on collaborative filtering | |
Luo et al. | Exploring destination image through online reviews: an augmented mining model using latent Dirichlet allocation combined with probabilistic hesitant fuzzy algorithm | |
CN104572915A (en) | User event relevance calculation method based on content environment enhancement | |
CN116860978B (en) | Primary school Chinese personalized learning system based on knowledge graph and large model | |
Jiang et al. | LDA-based online intelligent courses recommendation system | |
CN106484795A (en) | A kind of interest based on non-structured web page data recommends method | |
CN110209767A (en) | A kind of user's portrait construction method | |
Yu et al. | The personalized recommendation algorithms in educational application |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
C10 | Entry into substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
C14 | Grant of patent or utility model | ||
GR01 | Patent grant | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20170215 |
|
CF01 | Termination of patent right due to non-payment of annual fee |