CN106528693B - Educational resource recommended method and system towards individualized learning - Google Patents
Educational resource recommended method and system towards individualized learning Download PDFInfo
- Publication number
- CN106528693B CN106528693B CN201610942146.7A CN201610942146A CN106528693B CN 106528693 B CN106528693 B CN 106528693B CN 201610942146 A CN201610942146 A CN 201610942146A CN 106528693 B CN106528693 B CN 106528693B
- Authority
- CN
- China
- Prior art keywords
- learner
- educational resource
- learning
- resource
- educational
- 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
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
Abstract
Present invention relates particularly to educational resource recommended methods and system towards individualized learning, this method comprises: step S1, acquiring the essential characteristic data of learner, and establish according to the essential characteristic data resource interest-degree model and learning style model of learner;Wherein, the essential characteristic data include learner's identity information, learning behavior record and educational resource operation note;Step S2, learner characteristics model is established according to the resource interest-degree model and learning style model;Step S3, according to the learner characteristics model, the similarity between the similarity or different educational resource between different learners is calculated;Step S4, according to the similarity between the similarity or different educational resource between the different learners, the educational resource recommendation results for being directed to target learner are generated.The technical solution provided through the invention is, it can be achieved that be directed to the educational resource personalized recommendation of target learner.
Description
Technical field
The present invention relates to towards individualized learning educational resource recommend and big data studying technological domain, and in particular to face
To the educational resource recommended method and system of individualized learning.
Background technique
In recent years, universal with the ideas such as every citizen is committed to learning and pursues lifelong learning, Informal Learning has obtained more and more people
Concern.The development of mobile communication technology is so that the mode of Informal Learning experienced by digital studying to mobile learning again to such as
The development process of modern ubiquitous study, mode of learning become more flexible, convenient and personalized.Ubiquitous study can make to learn because of it
Habit person whenever and wherever possible using the characteristic of any terminal study, has adapted to demand of the modern society to study according to their own needs,
By the popular welcome of people, and become the development trend of the following study.And ubiquitous education resource is that ubiquitous study is carried out
Important support, therefore, how efficient, convenient, the personalized ubiquitous education resource of design and development become it is ubiquitous study grinds
The important content studied carefully.
At this stage, internet becomes the important channel that learner obtains education resource, however, with Internet resources quantity
Volatile growth, information is obtained only by traditional search engine mode, and there is return the result more, accuracy difference etc.
Disadvantage, this makes learner that can not obtain satisfied education resource.
The acquisition platform of education resource increases, and resource type also becomes more various, in face of the education resource of magnanimity, how
According to the action trail of learner, analytic learning person's interest is recommended the education resource for being able to reflect learner's learning interest, is realized
" teaching students in accordance with their aptitude " becomes more and more important, and the development that the individualized learning of network has become world today Web education field becomes
Gesture.And personalized recommendation technology in the successful application of commercial field provides a new think of to solve this problem in recent years
Road.Application of the personalized recommendation in internet electronic business field is highly developed, and researchers also start to explore one after another
Application of the recommended technology in education sector.Personalized recommendation technology is introduced into education sector, and is combined with ubiquitous education resource,
Learner can be helped to obtain personalized ubiquitous education resource well.
Summary of the invention
In view of this, it is an object of the invention to overcome the deficiencies of the prior art and provide the education towards individualized learning
Resource recommendation method and system realize the personalized recommendation of Network Educational Resources.
In order to achieve the above object, the present invention adopts the following technical scheme:
Educational resource recommended method towards individualized learning, comprising:
Step S1, the essential characteristic data of learner are acquired, and establish the money of learner according to the essential characteristic data
Source interest-degree model and learning style model;Wherein, the essential characteristic data include learner's identity information, learning behavior note
Record and educational resource operation note;
Step S2, learner characteristics model is established according to the resource interest-degree model and learning style model;
Step S3, according to the learner characteristics model, the similarity or different educational money between different learners are calculated
Similarity between source;
Step S4, it according to the similarity between the similarity or different educational resource between the different learners, generates
For the educational resource recommendation results of target learner.
Preferably, the study that learner logs in Educational Resources Website is stored in the step S1 in resource interest-degree model
Person's name and password, browsing object ID and the scoring of educational resource interest-degree;Wherein,
The educational resource interest-degree scoring intrScore is calculated according to formula (1):
IntrScore=Score/5+ (st+scc+sds)+Score (1)
Wherein, after Score is learner's use or browses educational resource, the scoring to educational resource, Score ∈ [1,5],
IntrScore ∈ [1,10];
St is that object ID browses duration score, is calculated according to formula (2):
Wherein, ToltalTime be use or browsing educational resource total time-consuming it is long, Mintime be the educational resource correspondence
Most short learning time, the unit of ToltalTime and Mintime are all minute;Click is number of clicks;
Scc is collection comment score, is calculated according to formula (3):
Wherein, Collect is collection number, and Comment is comment label number;
Sds is downloading recommendation scores, is calculated according to formula (4):
Wherein, Download is download time, and Share is to recommend good friend's number.
Preferably, the learning style deviation value of learner, are stored in the step S1 in learning style model
Style deviation value is practised to be calculated according to following steps:
Learner is divided into four kinds of learning styles and calculated every by the learning style assessment table filled in online according to learner
The deviation value tp of kind learning stylei, i=1,2,3,4;
According to the corresponding educational resource operation behavior classification value of four kinds of learning styles, it is biased to weight and deviation
Threshold value establishes learner's educational resource operation behavior classification and is biased to threshold values model;Wherein, the educational resource operation behavior classification
Value includes duration, clicking rate, time ratio and the number ratio of the corresponding educational resource operation behavior classification of every kind of learning style;
Every kind of learning style deviation value ty is calculated according to formula (5) and formula (6)i, i=1,2,3,4;
Wherein, ctypejFor the corresponding educational resource operation behavior classification value of i-th kind of learning style, j is positive integer, wjFor
ctypejCorresponding deviation weight, ljThreshold value, h are biased to for firstjThreshold value is biased to for second;
If the educational resource operation note number of learner is less than preset value, it is inclined that every kind of learning style is calculated with formula (7)
To value sti, otherwise, every kind of learning style deviation value st is calculated with formula (8)i;
sti=0.8*tpi+0.2tyi(7);sti=0.8*tyi+0.2tpi (8)。
Preferably, the learner characteristics model in the step S3 is the matrix information table of n* (m+4);The learner is special
N-th of learner is stored in sign model to m-th of the educational resource interest-degree scoring for using or browsing and n-th of learner
The deviation value of i-th kind of learning style;Total educational resource that 1≤n≤learner's total number of persons, 1≤m≤learner use or browsed
Number, i=1,2,3,4.
Preferably, the similarity in the step S3 between different learners is calculated according to following steps:
Count any learner UaUsing or the educational resource set S that browsedaWith any learner UbUsing or browsed
Educational resource set Sb, and seek SaAnd SbIntersection: Sab=Sa∩Sb;
Learner U is calculated according to formula (9)aAnd UbTo being averaged for the educational resource interest-degree scoring for using or browsing
Value:
Wherein, k=a, b, gk,mFor learner UkInterest-degree scoring to m-th of the educational resource for using or browsing, | Sk| it is
Learner UkUsing or the educational resource set S that browsedkIn element number, gk,xWith | Sk| all from the learner characteristics mould
Type is inquired to obtain;
Learner u is calculated according to formula (10)aAnd ubLearning style deviation value average value:
Wherein, fk,iFor learner UkThe deviation value of i-th kind of learning style;
Learner u is calculated according to formula (11)aAnd ubSimilarity:
Wherein, 0.7 and 0.3 is similarity score weight.
Preferably, it is generated according to the similarity between the different learners for target learner's in the step S4
Educational resource recommendation results, comprising:
According to Usim (ua,ub) calculated result, by Usim (ua,ub) descending sort;
Usim (u after taking descending to arrangea,ub) preceding K1It is a to be worth corresponding ubAs target learner uaReference learning person
Set:
Target learner u is predicted according to formula (12)aTo reference learning person ubUsing or m-th of educational resource browsing
Interest-degree scoring:
Wherein, gb,mFor reference learning person ubInterest-degree scoring to m-th of the educational resource for using or browsing, from
Habit person's characteristic model is inquired to obtain;For reference learning person ubTo being averaged for the educational resource interest-degree scoring for using or browsing
Value;
According to Pa,mCalculated result, by Pa,mDescending arrangement;
P after taking descending to arrangea,mThe corresponding educational resource of top n value recommend target learner ua, 1≤N≤m.
Preferably, the similarity in the step S3 between different educational resource is calculated according to following steps:
Statistics uses or browsed any educational resource raLearner setStatistics uses or browsed any education
Resource rbLearner setAnd it asksWithIntersection
Educational resource r is calculated separately according to formula (13)aAnd rbInterest-degree scoring average value:
Wherein, k=a, b,It is n-th of learner to educational resource rkInterest-degree scoring,To use or browsing
Educational resource rkLearner set in element number,WithAll obtained from the learner characteristics pattern query;
According to formula (14) computing education resource raAnd rbSimilarity:
Preferably, it generates and learns for target according to the similarity between the different educational resource in the step S4
The educational resource recommendation results of person, comprising:
According to Rsim (ra,rb) calculated result, by Rsim (ra,rb) descending sort;
Rsim (r after taking descending to arrangea,rb) preceding K2It is a to be worth corresponding rbAs Target Education resource raReference education
Resource collection:
Target learner u is predicted according to formula (15)aTo Target Education resource raInterest-degree scoring:
Wherein,For target learner uaTo educational resource rbInterest-degree scoring, from the learner characteristics pattern query
It obtains;
According toCalculated result, willDescending arrangement;
After taking descending to arrangeTop n be worth corresponding educational resource raRecommend target learner ua。
Preferably, the output result of the step S1~S4 is showed in the form of webpage chart.
A kind of educational resource recommender system towards individualized learning, comprising:
First model building module, for acquiring the essential characteristic data of learner, and according to the essential characteristic data
Establish the resource interest-degree model and learning style model of learner;Wherein, the essential characteristic data include learner's identity
Information, learning behavior record and educational resource operation note;
Second model building module, for establishing according to learner's resource interest-degree model and learning style model
Learner characteristics model;
Similarity calculation module, for calculating the similarity between different learners according to the learner characteristics model
Or the similarity between different educational resource;
Educational resource recommending module, for according between the similarity or different educational resource between the different learners
Similarity, generate be directed to target learner educational resource recommendation results.
The invention adopts the above technical scheme, at least have it is following the utility model has the advantages that
As shown from the above technical solution, this educational resource recommended method towards individualized learning provided by the invention,
Using cloud computing and big data technology, acquire the essential characteristic data of learner, establish learner resource interest-degree model and
Learning style model, and learner characteristics model is established according to resource interest-degree model and learning style model, to calculate difference
The similarity between similarity or different educational resource between learner, finally according between different learners similarity or
Similarity between different educational resource generates the educational resource recommendation results for being directed to target learner.It provides through the invention
Technical solution, be able to achieve the personalized recommendation of the Network Educational Resources for target learner.
In addition, the present invention is based on personalized recommendation technologies more mature in field of software engineering e-commerce, using association
Same filtering recommendation algorithms, and optimization is improved to the algorithm, make the personalized recommendation it is suitable for education resource.By this hair
Bright technical solution can improve learning efficiency, the learning quality of learner, improve teaching quality, learner is without access
Numerous and jumbled learning object repository blindly searches oneself interested educational resource, and learner is made conveniently and efficiently to obtain useful information,
It avoids learner and blindly searches education resource, while also saving the management cost of manager.
Furthermore the present invention makes full use of cloud computing and big data processing technique, the acquisition, storage, meter of mass data are realized
It calculates and effect of visualization is presented, can make full use of and save computing resource, there is good expansion and flexibility, to magnanimity knot
Structure and unstructured data carry out Quick Acquisition, processing and calculating, and the educational resource towards individualized learning can be rapidly completed
Various magnanimity calculates in recommender system.
Detailed description of the invention
Fig. 1 is the process signal for the educational resource recommended method towards individualized learning that one embodiment of the invention provides
Figure;
Fig. 2 is the schematic block diagram for the educational resource recommender system towards individualized learning that one embodiment of the invention provides.
Specific embodiment
Below by drawings and examples, technical scheme of the present invention will be described in further detail.
The educational resource recommended method towards individualized learning provided referring to Fig. 1, one embodiment of the invention, comprising:
Step S1, the essential characteristic data of learner are acquired, and establish the money of learner according to the essential characteristic data
Source interest-degree model and learning style model;Wherein, the essential characteristic data include learner's identity information, learning behavior note
Record and educational resource operation note;
Step S2, learner characteristics model is established according to the resource interest-degree model and learning style model;
Step S3, according to the learner characteristics model, the similarity or different educational money between different learners are calculated
Similarity between source;
Step S4, it according to the similarity between the similarity or different educational resource between the different learners, generates
For the educational resource recommendation results of target learner.
Wherein, learner's identity information storage format in step S1 is as shown in following table one:
Attribute-name | Data type | Explanation | Key |
StdID | Varchar | Learner ID Key Word | S_ID |
StdName | Varchar | Title-the pet name | S_Name |
Bithday | Date | Date of birth | S_B |
Grade | Varchar | Grade | S_G |
FavorEFG | Varchar | Hobby | S_F |
IP | Varchar | Learner uses IP | S_IP |
Table one
Learning behavior record storage format in step S1 is as shown in following table two:
Table two
Educational resource operation note storage format in step S1 is as shown in following table three:
Attribute-name | Data type | Explanation | Key |
StdID | Varchar | Learner ID Key Word | S_ID |
ObjectID | Varchar | Object ID | O_I |
Click | int | Hits, 0 (Def is not clicked on) | O_C |
Collect | int | Collection, 0 (Def is not collected) | O_Collect |
Download | int | Downloading time, 0 (Def is not downloaded) | O_Dl |
Comment | int | Comment label, 0 (Def is not commented on) | O_Comment |
Share | int | Recommend good friend's number, 0 (Def does not recommend) | O_Share |
TotalTime | int | It is long (minute) using the total time-consuming of/browsing resource | O_TotalTime |
Score | int | It scores 1-5 points, 0 (Def) | O_Score |
Table three
It should be noted that field " attribute-name ", " data type " and " explanation " in above-mentioned table one, table two and table three
Value is stored in relevant database MySQL, and the value of field " Key " is stored in distributed memory system HBase, by big data
It is written back in relevant database after processing platform statistics.
As shown from the above technical solution, this educational resource recommended method towards individualized learning provided by the invention,
Using cloud computing and big data technology, acquire the essential characteristic data of learner, establish learner resource interest-degree model and
Learning style model, and learner characteristics model is established according to resource interest-degree model and learning style model, to calculate difference
The similarity between similarity or different educational resource between learner, finally according between different learners similarity or
Similarity between different educational resource generates the educational resource recommendation results for being directed to target learner.It provides through the invention
Technical solution, be able to achieve the personalized recommendation of the Network Educational Resources for target learner.
Preferably, the study that learner logs in Educational Resources Website is stored in the step S1 in resource interest-degree model
Person's name and password, browsing object ID and the scoring of educational resource interest-degree;Data memory format in resource interest-degree model is as follows
Shown in table four:
Attribute-name | Data type | Explanation | Key |
StdID | Varchar | Learner ID KeyWord | S_ID |
ObjectID | Varchar | Object ID | O_I |
IntrScore | Float | The scoring of educational resource interest-degree | O_IntrScore |
Table four
Wherein, the educational resource interest-degree scoring intrScore in table four is calculated according to formula (1):
IntrScore=Score/5+ (st+scc+sds)+Score (1)
Wherein, after Score is learner's use or browses educational resource, the scoring to educational resource, Score ∈ [1,5],
IntrScore ∈ [1,10];
St is that object ID browses duration score, is calculated according to formula (2):
Wherein, ToltalTime be use or browsing educational resource total time-consuming it is long, Mintime be the educational resource correspondence
Most short learning time, the unit of ToltalTime and Mintime are all minute;Click is number of clicks;
Scc is collection comment score, is calculated according to formula (3):
Wherein, Collect is collection number, and Comment is comment label number;
Sds is downloading recommendation scores, is calculated according to formula (4):
Wherein, Download is download time, and Share is to recommend good friend's number.
Preferably, the learning style deviation value of learner, are stored in the step S1 in learning style model
Style deviation value is practised to be calculated according to following steps:
Learner is divided into four kinds by the learning style assessment table (referring to such as following table five) filled in online according to learner
It practises style and calculates the deviation value tp of every kind of learning stylei, i=1,2,3,4;
Table five
In above-mentioned table five, for each type, calculated result is scored at { 11A, 9A, 7A, 5A } → { 5,4,3,2 };
{ 3A, A, B, 3B } → { 1,0.5, -0.5, -1 };{ 5B, 7B, 9B, 11B } → { -2, -3, -4, -5 };When score ∈ { 5,4,3,2 }
Indicate that its learning style is biased to the former each type of i.e. ∈ { active type, perception type, optic type, sequence type }, when score ∈ 1,
0.5, -0.5, -1 } indicate that its learning style is biased to each type of compromise;When score ∈ { -2, -3, -4, -5 } ∈ reflective style,
It is Intuition, verbal type, comprehensive };The value of numerical value indicates the degree that its learning style is biased to.
According to the corresponding educational resource operation behavior classification value of four kinds of learning styles, it is biased to weight and deviation
Threshold value establishes learner's educational resource operation behavior classification and is biased to threshold values model (referring to table six);Wherein, the educational resource behaviour
Make the duration, clicking rate, time ratio that behavior classification value includes the corresponding educational resource operation behavior classification of every kind of learning style
With number ratio (referring to table seven);
Table six
Table seven
It should be noted that the value of the field " attribute-name ", " data type " and " explanation " in above-mentioned table seven is stored in pass
It is in type database MySQL, the value of field " Key " is stored in distributed memory system HBase, is united by big data processing platform
It is written back in relevant database after meter.
Every kind of learning style deviation value ty is calculated according to formula (5) and formula (6)i, i=1,2,3,4;
Such as: assuming that the ctype of the corresponding operation behavior classification value t_std of active type/reflective style1Value is 4.5, partially
To weight w1It is 0.3;The ctype of t_disc2Value is 3.8, is biased to weight w2It is 0.2;The ctype of n_disc3Value is 4.8,
It is biased to weight w3It is 0.25;The ctype of n_mes4Value is 4.2, is biased to weight w4It is 0.25.Then learner's active type/reflective style
The deviation value of learning style are as follows:
=4.5*0.3+3.8*0.2+4.8*0.25+4.2*0.25=4.36.
Ctype in formula (5)jIt is calculated according to formula (6):
Wherein, ctypejFor the corresponding educational resource operation behavior classification value of i-th kind of learning style, j is positive integer, wjFor
ctypejCorresponding deviation weight, ljThreshold value, h are biased to for firstjThreshold value is biased to for second;
Such as: set ctypejFor t_std, t_std known to inquiry table 7 is that learner participates in learning average duration, it is assumed that is
60 minutes, corresponding l1It is 30 minutes, h1It is 75 minutes, due to 60 ∈ (30,75), then the value of t_std are as follows:
If the educational resource operation note number of learner is less than preset value, it is inclined that every kind of learning style is calculated with formula (7)
To value sti, otherwise, every kind of learning style deviation value st is calculated with formula (8)i;
sti=0.8*tpi+0.2tyi(7);sti=0.8*tyi+0.2tpi (8)。
Preferably, the learner characteristics model in the step S3 is the matrix information table (referring to table eight) of n* (m+4);Institute
State be stored in learner characteristics model n-th of learner to use or browsed m-th of educational resource interest-degree scoring and
The deviation value of n-th of learner, i-th kind of learning style;1≤n≤learner's total number of persons, 1≤m≤learner use or browsed
Total educational resource number, i=1,2,3,4.
Table eight
Preferably, the similarity in the step S3 between different learners is calculated according to following steps:
Count any learner UaUsing or the educational resource set S that browsedaWith any learner UbUsing or browsed
Educational resource set Sb, and seek SaAnd SbIntersection: Sab=Sa∩Sb;
Learner U is calculated according to formula (9)aAnd UbTo being averaged for the educational resource interest-degree scoring for using or browsing
Value:
Wherein, k=a, b, gk,mFor learner UkInterest-degree scoring to m-th of the educational resource for using or browsing, | Sk| it is
Learner UkUsing or the educational resource set S that browsedkIn element number, gk,xWith | Sk| all from the learner characteristics mould
Type is inquired to obtain;
Learner u is calculated according to formula (10)aAnd ubLearning style deviation value average value:
Wherein, fk,iFor learner UkThe deviation value of i-th kind of learning style;
Learner u is calculated according to formula (11)aAnd ubSimilarity:
Wherein, 0.7 and 0.3 is similarity score weight.
Preferably, it is generated according to the similarity between the different learners for target learner's in the step S4
Educational resource recommendation results, comprising:
According to Usim (ua,ub) calculated result, by Usim (ua,ub) descending sort;
Usim (u after taking descending to arrangea,ub) preceding K1It is a to be worth corresponding ubAs target learner uaReference learning person
Set:
Target learner u is predicted according to formula (12)aTo reference learning person ubUsing or m-th of educational resource browsing
Interest-degree scoring:
Wherein, gb,mFor reference learning person ubInterest-degree scoring to m-th of the educational resource for using or browsing, from
Habit person's characteristic model is inquired to obtain;For reference learning person ubTo being averaged for the educational resource interest-degree scoring for using or browsing
Value;
According to Pa,mCalculated result, by Pa,mDescending arrangement;
P after taking descending to arrangea,mThe corresponding educational resource of top n value recommend target learner ua, 1≤N≤m.
Specifically, the P by MapReduce iterative process, after taking descending to arrangea,mTop n be worth corresponding education and provide
Recommend target learner u in sourcea。
Preferably, the similarity in the step S3 between different educational resource is calculated according to following steps:
Statistics uses or browsed any educational resource raLearner setStatistics uses or browsed any education
Resource rbLearner setAnd it asksWithIntersection
Educational resource r is calculated separately according to formula (13)aAnd rbInterest-degree scoring average value:
Wherein, k=a, b,It is n-th of learner to educational resource rkInterest-degree scoring,To use or browsing
Educational resource rkLearner set in element number,WithAll obtained from the learner characteristics pattern query;
According to formula (14) computing education resource raAnd rbSimilarity:
Preferably, it generates and learns for target according to the similarity between the different educational resource in the step S4
The educational resource recommendation results of person, comprising:
According to Rsim (ra,rb) calculated result, by Rsim (ra,rb) descending sort;
Rsim (r after taking descending to arrangea,rb) preceding K2It is a to be worth corresponding rbAs Target Education resource raReference education
Resource collection:
Target learner u is predicted according to formula (15)aTo Target Education resource raInterest-degree scoring:
Wherein,For target learner uaTo educational resource rbInterest-degree scoring, from the learner characteristics pattern query
It obtains;
According toCalculated result, willDescending arrangement;
After taking descending to arrangeTop n be worth corresponding educational resource raRecommend target learner ua。
Specifically, by MapReduce iterative process, after taking descending to arrangeTop n be worth corresponding education and provide
Source raRecommend target learner ua。
Preferably, above-mentioned educational resource is with educational resource recommendation list, the list of individualized education resource recommendation, learner society
The forms such as area's recommendation list recommend target learner.
Preferably, the output result of the step S1~S4 is showed in the form of webpage chart (such as passes through classification
The forms such as statistical chart, curve graph are presented in Web page).
Referring to fig. 2, the educational resource recommender system 100 towards individualized learning, comprising:
First model building module 101, for acquiring the essential characteristic data of learner, and according to the essential characteristic number
According to the resource interest-degree model and learning style model for establishing learner;Wherein, the essential characteristic data include learner's body
Part information, learning behavior record and educational resource operation note;
Second model building module 102, for building according to learner's resource interest-degree model and learning style model
Vertical learner characteristics model;
Similarity calculation module 103, for calculating similar between different learners according to the learner characteristics model
Similarity between degree or different educational resource;
Educational resource recommending module 104, for according to the similarity or different educational resource between the different learners
Between similarity, generate be directed to target learner educational resource recommendation results.
Preferably, the above-mentioned educational resource recommender system 100 towards individualized learning is disposed as follows:
1, cloud service is configured.Wisdom cooperative education Learning Service platform is installed, relational database MySQL is disposed in configuration,
JavaScript data, which are disposed, in each commercial page (Web, App etc.) on demand acquires script.
2, infrastructure equipment is arranged, software environment VMware workstation 8.04, Ubuntu is installed
12.04server editions, JDK1.7, Hadoop 2.6, Hive 0.7.0, Hbase0.90.3, Mahout0.9.0.
3, by statistical information used in data analysis process and similarity calculation and data, dispose Hadoop HDFS and
The MapReduce service code realizing analysis statistics and calculating.
4, acquire data and by tables of data be cut into blocks of files be stored in Distribute file system HDFS carry out storage and
Management.
5, data analysis task is executed.
6, data result is presented in Web page.
Above-described specific embodiment has carried out further the purpose of the present invention, technical scheme and beneficial effects
It is described in detail, it should be understood that being not intended to limit the present invention the foregoing is merely a specific embodiment of the invention
Protection scope, all within the spirits and principles of the present invention, any modification, equivalent substitution, improvement and etc. done should all include
Within protection scope of the present invention.Term " first ", " second " are used for description purposes only, and are not understood to indicate or imply
Relative importance.Term " multiple " refers to two or more, unless otherwise restricted clearly.
Claims (8)
1. the educational resource recommended method towards individualized learning characterized by comprising
Step S1, the essential characteristic data of learner are acquired, and emerging according to the resource that the essential characteristic data establish learner
Interesting degree model and learning style model;Wherein, the essential characteristic data include learner's identity information, learning behavior record and
Educational resource operation note;
Step S2, learner characteristics model is established according to the resource interest-degree model and learning style model;
Step S3, according to the learner characteristics model, calculate similarity between different learners or different educational resource it
Between similarity;
Step S4, according to the similarity between the similarity or different educational resource between the different learners, generation is directed to
The educational resource recommendation results of target learner;
Wherein, be stored in resource interest-degree model in the step S1 learner log in Educational Resources Website learner name and
Password, browsing object ID and the scoring of educational resource interest-degree;Wherein,
The educational resource interest-degree scoring intrScore is calculated according to formula (1):
IntrScore=Score/5+ (st+scc+sds)+Score (1)
Wherein, after Score is learner's use or browses educational resource, the scoring to educational resource, Score ∈ [1,5],
IntrScore ∈ [1,10];
St is that object ID browses duration score, is calculated according to formula (2):
Wherein, ToltalTime be use or browsing educational resource total time-consuming it is long, Mintime be the educational resource it is corresponding most
The unit of short learning time, ToltalTime and Mintime are all minute;Click is number of clicks;
Scc is collection comment score, is calculated according to formula (3):
Wherein, Collect is collection number, and Comment is comment label number;
Sds is downloading recommendation scores, is calculated according to formula (4):
Wherein, Download is download time, and Share is to recommend good friend's number.
2. the educational resource recommended method according to claim 1 towards individualized learning, which is characterized in that the step
The learning style deviation value of learner is stored in S1 in learning style model, the learning style deviation value is according to following steps
It is calculated:
The learning style assessment table filled in online according to learner, is divided into four kinds of learning styles for learner and calculates every kind
Practise the deviation value tp of stylei, i=1,2,3,4;
According to the corresponding educational resource operation behavior classification value of four kinds of learning styles, it is biased to weight and is biased to threshold
Value establishes learner's educational resource operation behavior classification and is biased to threshold values model;Wherein, the educational resource operation behavior classification takes
Value includes duration, clicking rate, time ratio and the number ratio of the corresponding educational resource operation behavior classification of every kind of learning style;
Every kind of learning style deviation value ty is calculated according to formula (5) and formula (6)i, i=1,2,3,4;
tyi=∑jctypej*wj (5)
Wherein, ctypejFor the corresponding educational resource operation behavior classification value of i-th kind of learning style, j is positive integer, wjFor
ctypejCorresponding deviation weight, ljThreshold value, h are biased to for firstjThreshold value is biased to for second;
If the educational resource operation note number of learner is less than preset value, every kind of learning style deviation value is calculated with formula (7)
sti, otherwise, every kind of learning style deviation value st is calculated with formula (8)i;
sti=0.8*tpi+0.2tyi(7);sti=0.8*tyi+0.2tpi (8)。
3. the educational resource recommended method according to claim 2 towards individualized learning, which is characterized in that the step
Learner characteristics model in S3 is the matrix information table of n* (m+4);N-th of study is stored in the learner characteristics model
Deviation value of the person to the scoring of m-th of educational resource interest-degree and n-th of learner, i-th kind of learning style that use or browsed;1
Total educational resource number that≤n≤learner's total number of persons, 1≤m≤learner use or browsed, i=1,2,3,4.
4. the educational resource recommended method according to claim 3 towards individualized learning, which is characterized in that the step
Similarity in S3 between different learners is calculated according to following steps:
Count any learner UaUsing or the educational resource set S that browsedaWith any learner UbUsing or the religion that browsed
Educate resource collection Sb, and seek SaAnd SbIntersection: Sab=Sa∩Sb;
Learner U is calculated according to formula (9)aAnd UbTo the average value for the educational resource interest-degree scoring for using or browsing:
Wherein, k=a, b, gk,mFor learner UkInterest-degree scoring to m-th of the educational resource for using or browsing, | Sk| it is
Learner UkUsing or the educational resource set S that browsedkIn element number, gk,mWith | Sk| all from the learner characteristics mould
Type is inquired to obtain;
Learner u is calculated according to formula (10)aAnd ubLearning style deviation value average value:
Wherein, fk,iFor learner UkThe deviation value of i-th kind of learning style;
Learner u is calculated according to formula (11)aAnd ubSimilarity:
Wherein, 0.7 and 0.3 is similarity score weight.
5. the educational resource recommended method according to claim 4 towards individualized learning, which is characterized in that the step
The educational resource recommendation results for being directed to target learner are generated in S4 according to the similarity between the different learners, comprising:
According to Usim (ua,ub) calculated result, by Usim (ua,ub) descending sort;
Usim (u after taking descending to arrangea,ub) preceding K1It is a to be worth corresponding ubAs target learner uaReference learning person collection
It closes:1≤K1< n;
Target learner u is predicted according to formula (12)aTo reference learning person ubUsing or m-th of educational resource browsing it is emerging
Interesting degree scoring:
Wherein, gb,mFor reference learning person ubInterest-degree scoring to m-th of the educational resource for using or browsing, from the study
Person's characteristic model is inquired to obtain;For reference learning person ubTo being averaged for the educational resource interest-degree scoring for using or browsing
Value;
According to Pa,mCalculated result, by Pa,mDescending arrangement;
P after taking descending to arrangea,mThe corresponding educational resource of top n value recommend target learner ua, 1≤N≤m.
6. the educational resource recommended method according to claim 1 towards individualized learning, which is characterized in that the step
Similarity in S3 between different educational resource is calculated according to following steps:
Statistics uses or browsed any educational resource raLearner setStatistics uses or browsed any educational resource
rbLearner setAnd it asksWithIntersection
Educational resource r is calculated separately according to formula (13)aAnd rbInterest-degree scoring average value:
Wherein, k=a, b,It is n-th of learner to educational resource rkInterest-degree scoring,To use or browsing religion
Educate resource rkLearner set in element number,WithAll obtained from the learner characteristics pattern query;
According to formula (14) computing education resource raAnd rbSimilarity:
7. the educational resource recommended method according to claim 6 towards individualized learning, which is characterized in that the step
According to the similarity between the different educational resource in S4, the educational resource recommendation results for being directed to target learner, packet are generated
It includes:
According to Rsim (ra,rb) calculated result, by Rsim (ra,rb) descending sort;
Rsim (r after taking descending to arrangea,rb) preceding K2It is a to be worth corresponding rbAs Target Education resource raReference educational resource
Set:1≤K2≤m;
Target learner u is predicted according to formula (15)aTo Target Education resource raInterest-degree scoring:
Wherein,For target learner uaTo educational resource rbInterest-degree scoring, obtained from the learner characteristics pattern query
It arrives;
According toCalculated result, willDescending arrangement;
After taking descending to arrangeTop n be worth corresponding educational resource raRecommend target learner ua。
8. described in any item educational resource recommended methods towards individualized learning, feature exist according to claim 1~7
In the output result of the step S1~S4 is showed in the form of webpage chart.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610942146.7A CN106528693B (en) | 2016-10-25 | 2016-10-25 | Educational resource recommended method and system towards individualized learning |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201610942146.7A CN106528693B (en) | 2016-10-25 | 2016-10-25 | Educational resource recommended method and system towards individualized learning |
Publications (2)
Publication Number | Publication Date |
---|---|
CN106528693A CN106528693A (en) | 2017-03-22 |
CN106528693B true CN106528693B (en) | 2019-07-30 |
Family
ID=58293157
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201610942146.7A Expired - Fee Related CN106528693B (en) | 2016-10-25 | 2016-10-25 | Educational resource recommended method and system towards individualized learning |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN106528693B (en) |
Families Citing this family (20)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107220916A (en) * | 2017-05-27 | 2017-09-29 | 邹杰 | Pushing learning resource method and device |
CN107221212B (en) * | 2017-05-27 | 2020-08-14 | 广州大洋教育科技股份有限公司 | Online learner cognitive style analysis method based on time sequence |
CN107392812A (en) * | 2017-09-01 | 2017-11-24 | 安徽教育网络出版有限公司 | A kind of education of middle and primary schools resource application service system based on semanteme |
CN110019851A (en) * | 2017-09-19 | 2019-07-16 | 香港智能金融科技有限公司 | A kind of the deal maker's teaching auto recommending method and system of personalization |
CN108875071B (en) * | 2018-07-05 | 2021-03-19 | 中北大学 | Learning resource recommendation method based on multi-view interest |
CN108959642A (en) * | 2018-07-27 | 2018-12-07 | 百度在线网络技术(北京)有限公司 | Method and apparatus for information to be written |
CN108985993A (en) * | 2018-08-21 | 2018-12-11 | 北京航空航天大学 | A kind of the network self-adapting control system and control method of on-line study |
CN109213863B (en) * | 2018-08-21 | 2022-05-24 | 北京航空航天大学 | Learning style-based adaptive recommendation method and system |
CN109388746A (en) * | 2018-09-04 | 2019-02-26 | 四川文轩教育科技有限公司 | A kind of education resource intelligent recommendation method based on learner model |
CN109299380B (en) * | 2018-10-30 | 2021-07-02 | 浙江工商大学 | Exercise personalized recommendation method based on multi-dimensional features in online education platform |
CN109299398B (en) * | 2018-12-07 | 2020-11-06 | 广东小天才科技有限公司 | Method for recommending text content and service equipment |
CN109582875B (en) * | 2018-12-17 | 2021-02-02 | 武汉泰乐奇信息科技有限公司 | Personalized recommendation method and system for online medical education resources |
CN109461064A (en) * | 2018-12-24 | 2019-03-12 | 税心懿 | A kind of educational information resource recommender system |
CN109933730A (en) * | 2019-03-29 | 2019-06-25 | 沈阳师范大学 | A kind of excellent class of the master of education of middle and primary schools recommends resource application method and system |
CN110188277B (en) * | 2019-05-31 | 2021-06-25 | 苏州百智通信息技术有限公司 | Resource recommendation method and device |
CN110929161B (en) * | 2019-12-02 | 2023-04-07 | 南京莱斯网信技术研究院有限公司 | Large-scale user-oriented personalized teaching resource recommendation method |
CN111401525A (en) * | 2020-03-20 | 2020-07-10 | 珠海读书郎网络教育有限公司 | Adaptive learning system and method based on deep learning |
CN112395324B (en) * | 2020-11-09 | 2021-05-25 | 艾迪安逊教育科技发展(广州)有限公司 | Big data storage system for online education platform |
CN112380335A (en) * | 2020-11-24 | 2021-02-19 | 中教云智数字科技有限公司 | Digital education resource recommendation system |
CN116628339B (en) * | 2023-06-09 | 2023-11-17 | 国信蓝桥教育科技股份有限公司 | Educational resource recommendation method and system based on artificial intelligence |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102737120A (en) * | 2012-06-01 | 2012-10-17 | 西安交通大学 | Personalized network learning resource recommendation method |
CN104750839A (en) * | 2015-04-03 | 2015-07-01 | 魅族科技(中国)有限公司 | Data recommendation method, terminal and server |
CN105045827A (en) * | 2015-06-29 | 2015-11-11 | 北京邮电大学 | Familiarity based information recommendation method and apparatus |
CN105138624A (en) * | 2015-08-14 | 2015-12-09 | 北京矩道优达网络科技有限公司 | Personalized recommendation method based on user data of on-line courses |
CN105786983A (en) * | 2016-02-15 | 2016-07-20 | 云南电网有限责任公司 | Employee individualized-learning recommendation method based on learning map and collaborative filtering |
-
2016
- 2016-10-25 CN CN201610942146.7A patent/CN106528693B/en not_active Expired - Fee Related
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102737120A (en) * | 2012-06-01 | 2012-10-17 | 西安交通大学 | Personalized network learning resource recommendation method |
CN104750839A (en) * | 2015-04-03 | 2015-07-01 | 魅族科技(中国)有限公司 | Data recommendation method, terminal and server |
CN105045827A (en) * | 2015-06-29 | 2015-11-11 | 北京邮电大学 | Familiarity based information recommendation method and apparatus |
CN105138624A (en) * | 2015-08-14 | 2015-12-09 | 北京矩道优达网络科技有限公司 | Personalized recommendation method based on user data of on-line courses |
CN105786983A (en) * | 2016-02-15 | 2016-07-20 | 云南电网有限责任公司 | Employee individualized-learning recommendation method based on learning map and collaborative filtering |
Also Published As
Publication number | Publication date |
---|---|
CN106528693A (en) | 2017-03-22 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN106528693B (en) | Educational resource recommended method and system towards individualized learning | |
US10841743B2 (en) | Branching mobile-device to system-namespace identifier mappings | |
CN107862553A (en) | Advertisement real-time recommendation method, device, terminal device and storage medium | |
US10776885B2 (en) | Mutually reinforcing ranking of social media accounts and contents | |
CN106296312A (en) | Online education resource recommendation system based on social media | |
CN110597962B (en) | Search result display method and device, medium and electronic equipment | |
CN101393566A (en) | Information tracking and detecting method and system based on network structure user pattern of behavior | |
CN112104642B (en) | Abnormal account number determination method and related device | |
CN102651011B (en) | Method and system for determining document characteristic and user characteristic | |
Kershaw et al. | Towards modelling language innovation acceptance in online social networks | |
CN105302880A (en) | Content correlation recommendation method and apparatus | |
CN108959580A (en) | A kind of optimization method and system of label data | |
CN104050243A (en) | Network searching method and system combined with searching and social contact | |
CN112839063A (en) | Message pushing method, message display method, storage medium and computer equipment | |
Kong et al. | Application Research of Personalized Recommendation Technology in College English Teaching Reform under The Background of Big Data | |
CN110717089A (en) | User behavior analysis system and method based on weblog | |
Hentschel et al. | Big data benefits for the software measurement community | |
Setiawan et al. | Virtual application technology of citizen journalism based on mobile user experience | |
Fitri et al. | Implementation of System Dynamics to Increase Successful Utilization of E-Commerce with Marketing and Innovation Capability in MSMEs | |
CN113468206A (en) | Data maintenance method, device, server, medium and product | |
CN108763400B (en) | Object dividing method and device based on object behaviors and theme preferences | |
Liu et al. | Personalized learning resource recommendation algorithm of mobile learning terminal | |
Qiao et al. | Research on web application technology in distance education personalized recommendation system | |
Jia et al. | A mental health education strategy based on online videos | |
Lin et al. | Machine Learning-based Weibo user group profiling under hot events |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
C06 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant | ||
CF01 | Termination of patent right due to non-payment of annual fee | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20190730 Termination date: 20201025 |