CN104866557B - A kind of personalized instant learning theoretical based on constructive learning supports System and method for - Google Patents
A kind of personalized instant learning theoretical based on constructive learning supports System and method for Download PDFInfo
- Publication number
- CN104866557B CN104866557B CN201510252479.2A CN201510252479A CN104866557B CN 104866557 B CN104866557 B CN 104866557B CN 201510252479 A CN201510252479 A CN 201510252479A CN 104866557 B CN104866557 B CN 104866557B
- Authority
- CN
- China
- Prior art keywords
- knowledge
- user
- micro
- relation
- point
- 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 29
- 238000004458 analytical method Methods 0.000 claims abstract description 10
- 238000010276 construction Methods 0.000 claims abstract description 10
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 8
- 230000006870 function Effects 0.000 claims description 6
- 230000011218 segmentation Effects 0.000 claims description 6
- 238000003860 storage Methods 0.000 claims description 5
- 230000032798 delamination Effects 0.000 claims description 3
- 230000007717 exclusion Effects 0.000 claims description 3
- 239000000203 mixture Substances 0.000 claims description 3
- 241000270322 Lepidosauria Species 0.000 claims description 2
- 230000007547 defect Effects 0.000 claims description 2
- 230000010354 integration Effects 0.000 claims description 2
- 238000012216 screening Methods 0.000 claims description 2
- 238000000605 extraction Methods 0.000 claims 2
- 239000004744 fabric Substances 0.000 claims 1
- 230000000875 corresponding effect Effects 0.000 description 9
- 239000008186 active pharmaceutical agent Substances 0.000 description 4
- 230000007812 deficiency Effects 0.000 description 4
- 238000004364 calculation method Methods 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 244000097202 Rathbunia alamosensis Species 0.000 description 2
- 235000009776 Rathbunia alamosensis Nutrition 0.000 description 2
- 238000001914 filtration Methods 0.000 description 2
- BYACHAOCSIPLCM-UHFFFAOYSA-N 2-[2-[bis(2-hydroxyethyl)amino]ethyl-(2-hydroxyethyl)amino]ethanol Chemical compound OCCN(CCO)CCN(CCO)CCO BYACHAOCSIPLCM-UHFFFAOYSA-N 0.000 description 1
- PEDCQBHIVMGVHV-UHFFFAOYSA-N Glycerine Chemical compound OCC(O)CO PEDCQBHIVMGVHV-UHFFFAOYSA-N 0.000 description 1
- 101100261006 Salmonella typhi topB gene Proteins 0.000 description 1
- 230000002596 correlated effect Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 235000013399 edible fruits Nutrition 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 238000010249 in-situ analysis Methods 0.000 description 1
- 239000003550 marker Substances 0.000 description 1
- 230000000644 propagated effect Effects 0.000 description 1
- 230000003252 repetitive effect Effects 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 238000005728 strengthening Methods 0.000 description 1
- 101150032437 top-3 gene Proteins 0.000 description 1
- 230000003313 weakening effect 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
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/36—Creation of semantic tools, e.g. ontology or thesauri
- G06F16/367—Ontology
-
- 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
- G06F16/958—Organisation or management of web site content, e.g. publishing, maintaining pages or automatic linking
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Databases & Information Systems (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Animal Behavior & Ethology (AREA)
- Computational Linguistics (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
System and method for is supported the invention provides a kind of personalized instant learning theoretical based on constructive learning, including the module such as the analysis of user knowledge demand action and record, the micro- knowledge base of label graph of a relation, micro- knowledge recommendation algorithm based on constructive learning.System realizes the construction demand that its knowledge is analyzed from personal user's social media information, is immediately then the personalized micro- knowledge of its recommendation study, so as to actively rapidly help user to make up lacking in knowledge.The method of the invention can help user to lift personal knowledge attainment intelligent and high-efficiency, obtain good economic and social benefit as a critical services content of internet social intercourse system.
Description
Technical field
The invention discloses user knowledge demand analysis in a kind of social networks, system that personalized instant knowledge is recommended with
Method, belong to the field of individualized knowledge recommendation.
Background technology
Nowadays, people's daily life increasingly be unable to do without social networks.It is single that Sina weibo data are shown, cut-off 2014
Year October, the microblogging moon active users reached 1.67 hundred million people.People realize that information is propagated, information is divided by social networks
Enjoy, acquisition of information.Illustrated herein using Sina weibo this social networks as platform.
Microblogging function is become stronger day by day, and the micro-blog information largely increased also result in information while bringing and facilitating
The problems such as overload.The different knowledge of each ken are contained in the microblog data that user receives daily, while comprising a large amount of
The garbage not theed least concerned with academic knowledge, user is caused to need active filtering largely irrelevant when " brush microblogging "
Information, useful information is screened to read, considerably increase burden during user's brush microblogging.
In addition, for the strange knowledge that some had not been contacted, user is examined by inputting keyword in a search engine
Suo Xiangguan webpage, valuable webpage is then therefrom selected to read.Therefore user, which often learns a new knowledge, will all undergo
Cumbersome retrieval screening process, and this process does not ensure that the content that user finally retrieves is that oneself is interested.
Therefore user's " brush microblogging " is one and had a glance at that planless learning process, the planning of neither one system, this may lead
Apply family repetitive learning, spend the unnecessary time.
Microblog users how are helped to screen effective information, filtering useless, uninterested information, and on this basis
The hobby of user is analyzed, the study that personalization is made to user is recommended, and is a urgent problem to be solved.
Study commending system solves posed problems above to a certain extent.But existing study commending system is also deposited
In place of some shortcomings：1. it is unable to the learning interest of user in real.2. content recommendation user's interest level is not high.3. learn
It is complicated to practise content-form, causes learning burden weight, reduces user's learning interest.4. no comprehensively learn to advise for custom system
Draw.
The deficiency for more than, the invention discloses a kind of personalized instant learning theoretical based on constructive learning to support system
With method.
The content of the invention
It is contemplated that making up the deficiency of above-mentioned study commending system, it is provided knows in a kind of social networks according to user
Knowledge demand recommends the study commending system of individualized learning resource to user immediately.The present invention focuses on builds property one by one for user
The academic environment of change, the instant discovery learning point of interest of user is helped, makes up lacking in knowledge, the complete structure of knowledge of construction.This hair
It is bright to consider emphatically at 2 points：1. in-situ analysis user knowledge demand, the desired knowledge missing of user is found, recommends study immediately
Resource.2. by introducing oriented label graph of a relation, realize that the education resource based on construction is recommended.
Personalized instant learning provided by the invention based on micro- Knowledge Construction supports System and method for, including following step
Suddenly：
User knowledge context analyzer：First by obtaining the microblog label of a large amount of microblogging certification users, statistical analysis difference
Then crowd predicts the affiliated crowd of the user by analyzing the microblog label of targeted customer, entered to the hobby of different microblog labels
And predict that targeted customer is interested in which ken.Then batch obtains user and its concern newest microblogging delivered of good friend
State, is integrated into a microblogging state document by the time order and function delivered, and document is automatically extracted with improved TF-IDF algorithms
Keyword.The keyword of one group of Weight is finally given, represents the knowledge back of the body of the knowledge point that user currently pays close attention to, as user
Scape.
Build oriented label graph of a relation：Oriented label graph of a relation stores existing a kind of inherent connection between all knowledge points
The relation such as system, i.e., including on knowledge content, mutual exclusion.The structure of knowledge of whole label graph of a relation is a netted structure.Mark
The content that label graph of a relation includes can be divided into several big kens, and the structure of knowledge of each ken is tree-like knot
Structure.Each big ken eventually forms a tree structure by the root node down subdivision from level to level of a knowledge.
Structure extension knowledge base：It is in order to meet further demand of the user to knowledge, when user is to being to extend knowledge base
When the micro- knowledge recommended of uniting is interested, system will recommend the extension knowledge for micro- knowledge to user, this be probably one with
The related website of micro- knowledge, or an article.
It was found that the desired missing knowledge of user：The knowledge that user acquires while brush microblogging be it is scattered incomplete,
The missing of knowledge be present.Oriented label graph of a relation provides a complete knowledge structure system, is formulated equivalent to for user
One comprehensive study plan.The knowledge point that user is currently paid close attention to is marked in label graph of a relation, therefrom discovery and user
Currently what concern knowledge point was closely related is still the knowledge that user had not contacted, user's study is recommended, so as to help to use
Family reaches the purpose that knowledge leakage detection is filled a vacancy.
Realize that the education resource based on construction is recommended:Using the desired missing knowledge of user as knowledge to be recommended, recommend
User learns.The process that user learns missing knowledge is exactly the process of a structure knowledge.User often learns a missing knowledge,
The structure of knowledge of user will be more perfect.
Brief description of the drawings
Method flow diagram is recommended in Fig. 1 study；
Fig. 2 user knowledge context analyzer flow charts
The affiliated crowd of Fig. 3 user predicts flow chart
Fig. 4 user's gender prediction's flow chart
Fig. 5 obtains targeted customer's microblog label flow chart
Fig. 6 age of user predicts flow chart
Fig. 7 user's occupation prediction flow chart
Fig. 8 user currently pays close attention to knowledge analysis flow chart
Fig. 9 label graphs of a relation
The tree-like branch of Figure 10 label graphs of a relation
Figure 11 creates micro- knowledge base flow chart
Figure 12 has found the knowledge process figure of missing
Figure 13 system architecture diagrams
Figure 14 working-flow figures
Embodiment
Each step in the present invention is described in detail with reference to the accompanying drawings.
As shown in figure 1, the invention provides a kind of personalized instant learning theoretical based on constructive learning to support method, bag
Include following steps：
User knowledge context analyzer：Use is analyzed by predicting the affiliated crowd of user and finding the knowledge currently paid close attention to of user
Family knowledge background.As shown in Fig. 2 user knowledge context analyzer comprises the following steps：
The affiliated crowd's prediction of user：The affiliated crowd of user is predicted by microblog label with Bayes' theorem.Such as Fig. 3 institutes
Show, the affiliated crowd's prediction of user comprises the following steps：
User gender prediction：User's sex is predicted by microblog label.As shown in figure 4, user gender prediction include with
Lower step：
Obtain the sex and label of a large amount of microblogging certification users：The API provided by microblogging open platform obtains a large amount of micro-
The information (including user's sex, age, occupation, microblog label, the microblogging state delivered) of rich certification user.
Label is counted by sex：The microblog users label of all acquisitions is divided by masculinity and femininity, male is obtained and likes
Microblog label storehouse and the microblog label storehouse liked of women.
Expandtabs storehouse：If some keyword frequently occurs in the microblogging state of user, it is considered that should
Keyword is the microblog label of the user.The microblogging state that all male users were delivered is counted, by microblogging denoising, is segmented,
Word frequency is counted, is added to the microblog label storehouse that male likes using the higher keyword of word frequency as the microblog label that male likes.
The microblogging state that all female users were delivered is counted, is liked by the use of same method using the higher keyword of word frequency as women micro-
Rich label is added to the microblogging mark storehouse that women likes.
Calculate the probability of each label appearance：The microblogging that the microblog label storehouse and women that statistics of male is liked respectively are liked
The probability that each label occurs in tag library.Assuming that a total of Tm label in microblog label storehouse that male likes, women like
A total of Tw label in microblog label storehouse, wherein label X occurs Sm times in the microblog label storehouse that male likes, is liked in women
Occurring Sw times in the microblog label storehouse of love, then the probability that label X occurs in the microblog label storehouse that male likes is Sm/Tm,
The probability occurred in the microblog label storehouse that women likes is Sw/Tw.
Obtain the microblog label of targeted customer：Pass through the microblog label of user's self marker, the microblogging of user good friend mark
The microblogging that label, user are delivered and forwarded finds the label that user likes.As shown in figure 5, obtain targeted customer's microblog label bag
Include following steps：
Directly obtain user's microblog label：The API provided by microblogging open platform obtains the microblog label of targeted customer.
Good friend is paid close attention to by user and obtains microblog label：Concern good friend is that user actively selects, normal conditions and target
User has similar interest hobby.The microblog label of concern good friend is obtained, the higher microblog label of the statistics frequency of occurrences is as mesh
The microblog label that mark user likes.
The microblogging status discovery microblog label delivered by user：User is delivered in the state of forwarding if frequently occurring one
A little keywords, then it is the microblog label that user likes to think these keywords frequently occurred.
Targeted customer gender prediction：The prediction possible sex of targeted customer is analyzed with Bayes' theorem.Assuming that Li is mesh
The microblog label that mark user likes, then according to Bayes' theorem, it is male (female that can calculate targeted customer when label Li occurs
Property) probability.It is assumed that masculinity and femininity likes that the probability of some label is 50%, formula is as follows：
Above formula is based on a label and analyzes the probability that targeted customer is male (women).One microblog users of ordinary circumstance
The microblog label liked has several, considers each microblog label that targeted customer likes, final probability calculation is public
Formula uses following formula：(setting Pi=P (Men | Li))
Analyzed more than, it is final obtain targeted customer be male user and be female user probability.Compare two generally
The size of rate, if the probability that targeted customer is male user is more than the probability that targeted customer is women, then it is assumed that targeted customer
It is on the contrary for male, then it is assumed that targeted customer is women.
Age of user is predicted：Age of user prediction flow chart is illustrated in figure 6, specific steps are referring to user gender prediction.
User's occupation prediction：User's occupation prediction flow chart is illustrated in figure 7, specific steps are referring to user gender prediction.
Know user knowledge interest preference from the affiliated crowd's prediction result of user.
User currently pays close attention to knowledge point analysis：The API provided by microblogging open platform obtains user and its concern good friend
The newest microblogging state delivered, therefrom analyze the knowledge point that user currently pays close attention to.As shown in the figure 8, user currently pays close attention to knowledge point
Analysis comprises the following steps：
Obtain user and its concern newest microblogging state delivered of good friend：The API provided by microblogging open platform is obtained
User and its concern newest microblogging state delivered of good friend, and microblogging state is incorporated into a text according to the time order and function delivered
In shelves.
Microblog data denoising：The invalid informations such as emoticon, link, the pet name, topic in removal microblog data.
Microblogging segments：The ICT segmenter provided with the Chinese Academy of Sciences segments to microblogging document.
Word segmentation result denoising：The invalid informations such as the adjective in word segmentation result, adverbial word, stop words are removed, obtaining one group has
The keyword (the corresponding correlated knowledge point of each effective keyword) of effect.
Key word analysis：The weight of each keyword is calculated with TF-IDF Algorithm Analysis.To i-th of pass in document
Keyword weight WiCalculation formula such as following formula：
Wherein tfiRepresent the word frequency that i-th of keyword occurs in document, max tfiRepeat in expression document most
Keyword, N represent the total number of documents of corpus, niThe number of files for including the keyword in corpus is represented, D represents entire chapter document
Length, diRepresent the length for the positional distance document beginning that i-th of keyword occurs first.The thought of TF-IDF algorithms includes
2 points：First point, the word frequency that keyword occurs is higher, and its weight is bigger；Second point, keyword is more common, and its weight is smaller.Such as
One keyword of fruit seldom occurs in other documents, and is frequently occurred in the document, then represents this keyword and be good at instead
Reflect the feature of the document.On the basis of original TF-IDF algorithms, the characteristics of combining microblogging state of the invention, considers each pass
Keyword delivers corresponding time order and function, delivers time shorter keyword, and its weight is bigger.Deliver the length of time in a document
Show as the distance length that the keyword starts to document.
Keyword weight is adjusted by user interest preference：Knowledge is adjusted by the user knowledge interest preference known in advance
Point weight.Each knowledge point judges whether it is knowledge point that the affiliated crowd of the user likes, if it is, strengthening the knowledge point
Weight, if it is not, then weakening the knowledge point weight.For example, the keyword that " basketball ", which is male user, more to be liked, if target
User is male, then strengthens the weight of the keyword with corresponding probability, if targeted customer is women, with corresponding probability
Weaken the weight of keyword.The final weight calculation formula of keyword is as follows：
Pi represents that crowd where targeted customer selects the probability of the keyword, and Pj represents that different crowd selects the keyword
Probability.Last represents that targeted customer selects being averaged for the keyword to the select probability of the keyword with crowd on the right of equation
The comparison of probability.
The keyword of one group of Weight is finally obtained by above step, each keyword corresponds to a knowledge point, this
A little knowledge points are the knowledge point that user currently most pays close attention to.
Label graph of a relation and micro- knowledge base：Label graph of a relation and micro- knowledge base need manual creation, and label graph of a relation is realized
The structure of knowledgeization stores and provides knowledge relation, and micro- knowledge base is used for the micro- knowledge entry for storing manual creation.Label graph of a relation
Comprise the following steps with micro- knowledge base：
Build oriented label graph of a relation：Exist between knowledge comprising the relation with mutual exclusion, by this knowledge relation and correspondingly
Knowledge point be stored in label graph of a relation.As shown in figure 9, whole label graph of a relation contains all knowledge points, and respectively
It is existing between knowledge point to include and mutex relation.
Knowledge classification：Substantial amounts of keyword is captured from network using reptile, by keyword by affiliated ken point
Into several classes, merge synonymous keyword.As shown in figure 9, the dashed circle of outer ring five in figure, each circle represents a knowledge neck
Domain.
Knowledge delamination：As shown in Figure 10, the keyword of each ken is showed in the form of tree structure.It is different
Keyword covering the scope of one's knowledge scope it is different.The keyword of root node (first layer) covers the knowledge of whole ken
Face；The ken is subdivided into several knowledge blocks by the keyword for the child node (second layer) inherited by root node；From different
Each knowledge blocks is divided into a series of more detailed knowledge points (third layer) by the child node (keyword) that knowledge blocks are inherited；According to
It is secondary down to be segmented by layer, until knowledge point can not subdivide, thus obtain one have a level direction can represent a certain knowledge neck
The crucial phrase of the tree structure in domain.As shown in figure 9, two thick dashed line circles represent knowledge delamination, it is as can be seen from the figure outer
The knowledge of layer is the inheritance of knowledge from internal layer.
The structure of knowledgeization stores：Each node represents a knowledge point, and the subordinate and coordination between node are corresponding
It is existing between knowledge point to include and mutex relation.Realize the structured storage of knowledge.As shown in figure 9, each solid line in figure
Small circle all represent a knowledge point, the line between small circle represents existing contact between knowledge point.
Label graph of a relation automatically updates：Knowledge point and the structure of knowledge in label graph of a relation in use can be continuous
Automatically update improvement.User has found that knowledge point expression inaccuracy can carry out edit-modify in use, if it find that having
The knowledge point user of omission can also create an entry according to given form (content of title+300-500 words) oneself and carry
Give system.
Create micro- knowledge base：One micro- knowledge of the corresponding establishment in each knowledge point in label graph of a relation, forms micro- knowledge
Storehouse.As shown in figure 11, micro- knowledge base is created to comprise the following steps：
The micro- knowledge of search engine retrieving：Search engine is inputted using the title of micro- knowledge as keyword, from Baidupedia, dimension
Base encyclopaedia, Chinese encyclopaedia net obtain the knowledge entry related to micro- knowledge.
Integrated searching result：The retrieval result of acquisition is integrated, merges the content of repetition, each knowledge point is integrated
Into the brief content of 300 word or so.
Store micro- knowledge：For each micro- knowledge using search key as title, the content of integration is knowledge body, is stored in micro-
Knowledge base.
It was found that the knowledge of missing：User is helped to find the knowledge of missing by the knowledge point of user's current interest, and
Recommend corresponding knowledge point for user, make up the deficiency of knowledge.The knowledge point that the representation of knowledge user of missing can not actively discover,
But this partial knowledge is that the knowledge point paid close attention at present with user is closely related, it is most likely that is the knowledge point needed for user.Such as
Shown in Figure 12, it is found that the knowledge of missing comprises the following steps：
User knowledge background is marked in label graph of a relation：User context knowledge (knowledge point currently most paid close attention to) is being marked
It is marked in label graph of a relation.
It was found that the knowledge of missing：Existing relation between the knowledge point provided by label graph of a relation, find to work as with user
The knowledge point of preceding concern other knowledge points in close relations, the as desired missing knowledge of user.
Structure and recommendation：The present invention helps user to find desired missing knowledge on the basis of original knowledge background, leads to
Cross the knowledge for constantly recommending study missing, the new knowledge hierarchy of construction.Construction comprises the following steps with recommendation：
Determine knowledge to be recommended：The knowledge point of each user concern can find a corresponding missing knowledge, will use
The desired missing knowledge top3 in family is as knowledge to be recommended.
Determine extension knowledge to be recommended：Fed back according to the study of user's last time login system, if user is to recommendation
Some knowledge point represents interested, then using the extension knowledge of the knowledge point as extension knowledge to be recommended.
Recommend knowledge：Knowledge to be recommended and extension knowledge recommendation to be recommended are learnt to user.
Construct knowledge：By helping user constantly to find to lack knowledge, user's study is recommended, progressively helps user more
Knowledge deficiency is mended, finally realizes Knowledge Construction.
The invention also discloses a kind of personalized instant learning theoretical based on constructive learning to support system, including：
It is as shown in figure 13 system architecture diagram, illustrates the composition of each module of the system, is worked as shown in figure 14 for system
Flow chart, illustrate whole flow process during system work.
The system is by interface module, proposed algorithm module, database design module three parts composition.
Interface module includes log-in interface, knowledge learning interface, study feedback interface.
Log-in interface：User logs in the system by social platform existing account, and system is worked as by account name unique mark
Before the user that logs in.
Knowledge learning interface：The possible micro- knowledge interested of three users of display, recommends user's study.
Learn feedback interface：User can be fed back after the knowledge of recommendation has been learnt, and the knowledge of recommendation is evaluated,
User's system in next login system recommends corresponding extension knowledge according to the conventional user that feeds back to of user.
Proposed algorithm module includes user knowledge context analyzer, finds missing knowledge, recommendation and construction three parts.
User knowledge context analyzer：System obtains social networks mandate, extracts user social contact information, predicts the affiliated people of user
Group, the knowledge point that analysis user pays close attention to instantly, so as to accurately understand the knowledge background of user.
It was found that missing knowledge：Knowledge background using the knowledge that user currently pays close attention to as user, sent out using label graph of a relation
The knowledge point that current family may lack.
Recommend education resource：Each login system recommends micro- knowledge of three missings to user.
Database design module includes label graph of a relation, micro- knowledge base, extension knowledge base three parts.
Label graph of a relation：All knowledge points, and existing contact between knowledge point are stored in label graph of a relation.Record
The title of each knowledge point, inherit from which knowledge point, which knowledge point derived from.
Micro- knowledge base：All knowledge points are stored, each knowledge point stores its title and brief knowledge content.It is micro- to know
The knowledge title known in knowledge title and label graph of a relation all in storehouse corresponds.
Extend knowledge base：All extension knowledge of knowledge library storage is extended, when user represents that sense is emerging to some knowledge point
When interesting, system can be directed to the extension knowledge of the knowledge point when user logs in next time to user's recommendation.Extension knowledge form includes
Detailed review, related web site, related forum, pertinent texts etc..
The present invention proposes that a kind of personalized instant learning theoretical based on constructive learning supports System and method for, from user's
User's knowledge interested is analyzed in social information, and in this, as the knowledge background of user, is provided with label graph of a relation
Contact between knowledge, help user to find the knowledge of missing, recommended while user's brush microblogging in the form of micro- knowledge
User learns, and has provided the user the knowledge services of personalization, helps user to eliminate the time of retrieval knowledge, compensate for knowledge
The defects of.
Actual use
Method described in the invention can as the One function module of social networks in internet, can also be used as with
The core implementation method of study website based on text, it is intended to recommend personalized learning knowledge to user, guiding user has found
The learning interest of oneself.
Claims (8)
1. a kind of personalized instant learning theoretical based on constructive learning supports method, it is characterised by：By analyzing user network
Social information, extraction can characterize the keyword background knowledge point of user's one group of Weight interested, currently be closed as user
Note knowledge point；The oriented label graph of a relation of a network structure is introduced on this basis, between label graph of a relation stored knowledge content
Include, mutex relation；User is currently paid close attention to knowledge point to be tagged in label graph of a relation, passes through extraction and current concern knowledge
Point has direct relation, but the relevant knowledge that user had not contacted recommends user's study as the knowledge that may be lacked；By not
Disconnectedly recommend the defects of potential missing knowledge for user to make up knowledge to user, help user's one complete knowledge of construction
Structural system；System is directed to each knowledge point one knowledge title of corresponding establishment and a 300 words left side in label graph of a relation
Right micro- knowledge content of text, all micro- knowledge combine the micro- knowledge base to form that property instant learning is supported one by one.
2. a kind of personalized instant learning theoretical based on constructive learning supports method, comprise the following steps：
User knowledge context analyzer：Obtain user and its pay close attention to one group of microblog data of the newest issue of good friend, therefrom analyze user
Current knowledge point of interest, and in this, as the knowledge background of user；
Build label graph of a relation：Label graph of a relation is made up of two parts：Relation between knowledge concepts and knowledge；Knowledge concepts：
The knowledge that whole label graph of a relation includes can be divided into the knowledge of several major classes, and a major class represents a ken；Often
The knowledge of one major class can be divided into different knowledge blocks again, and the knowledge of each block is segmented into different knowledge points again,
Subdivision is gone down always, until that can not subdivide；One tag representation in each knowledge point；Relation：Whole label graph of a relation by
Some tree structure compositions, each tree structure represent existing pass between all knowledge points that a ken is included
System；All tree structures are grouped together into a netted label graph of a relation；Each node of tree represents one and known
Know point；Sub- knowledge point is the subdivision of father knowledge point, and they are inclusion relations；The knowledge point of same layer is that mutual exclusion is closed in one tree
System；Thus a complete knowledge structure system is built；
Build micro- knowledge base：Each label occurred in label graph of a relation represents a micro- knowledge；Micro- knowledge base stores institute
There is the knowledge point included in label graph of a relation；Knowledge content of each micro- knowledge package containing a title and 300 words or so；Label
The title of each label and micro- knowledge in graph of a relation corresponds；The advantages of micro- knowledge, is that user can save study
Time, the out of patience content for going to read big length in detail of user, micro- knowledge can guarantee that user exists especially when brush microblogging
Clearly the knowledge point is oneself knowledge interested in short time；
Structure extension knowledge base：Knowledge content in extension knowledge base is extension to micro- knowledge, using search engine retrieving with
The related website of micro- knowledge, forum, article, blog Internet resources, the extension knowledge as micro- knowledge；
It was found that the desired missing knowledge of user：User's knowledge entry interested is marked in label graph of a relation, therefrom found
Closely related other knowledge entries, the knowledge that as user may lack with user's knowledge entry interested；
Knowledge recommendation：The knowledge grasped according to user, with reference to knowledge structure system, recommend the missing of personal expectation for user
Knowledge.
3. supporting method based on the theoretical personalized instant learning of constructive learning as described in claim 2, wherein user knows
Knowing context analyzer step includes：
(1) the affiliated crowd's prediction of user：The information of a large amount of microblogging certification users is captured, training in advance knows different crowd to various
Label is liked；Targeted customer's microblog label is obtained, analyzes which class crowd the user belongs to；Different crowds are to different field
Knowledge Interest Measure is different；
(2) batch obtains user and its pays close attention to the microblog data of the newest issue of good friend：Obtain user and its concern newest hair of good friend
One group of microblog data of cloth, the time order and function delivered according to state are integrated into a microblogging state document；
(3) microblog data denoising：Remove the noise of microblog data currently got, including emoticon, link, the pet name these not
Embody the character of user interest；
(4) microblogging segments：The microblogging state document Jing Guo denoising is subjected to word segmentation processing using segmentation methods；
(5) word segmentation result denoising：Remove the noise in word segmentation result；
(6) key word analysis：The weight of keyword is calculated with TF-IDF algorithms, introduces the position that keyword occurs on this basis
Put, crowd where targeted customer to the fancy grade of the keyword, finally provide one group of keyword for carrying weight；
(7) user knowledge background modeling：Sorted by weight size to keyword, first three great keyword of weighting, represent to use
The knowledge point that family is most paid close attention at present, as user knowledge background.
4. method is supported based on the theoretical personalized instant learning of constructive learning as described in claim 2, wherein being built with
Include to label graph of a relation step：
(1) knowledge classification：A number of keyword is captured from network using reptile, keyword is pressed into affiliated ken
Classification, and merge synonymous keyword；
(2) knowledge delamination：The keyword of each ken is showed in the form of tree structure；Different keyword coverings
The scope of one's knowledge scope it is different；The keyword of root node covers the scope of one's knowledge of whole ken；The son inherited by root node is tied
The ken is subdivided into several knowledge blocks by the keyword of point；From the child node that different knowledge blocks are inherited by each knowledge
Block is divided into a series of more detailed knowledge points；Down segmented by layer successively, until knowledge point can not subdivide, thus obtain one
It is individual having a level direction and the knowledge frame of relation between each knowledge point in a certain ken can be represented；
(3) structure of knowledgeization stores：Each node represents a knowledge point, and the subordinate and coordination between node are correspondingly known
Know existing between point include and mutex relation；Realize the structured storage of knowledge；
(4) label graph of a relation automatically updates：Knowledge point and the structure of knowledge in label graph of a relation in use can be continuous
Automatically update improvement.
5. method is supported based on the theoretical personalized instant learning of constructive learning as described in claim 2, wherein building micro-
Knowledge base step includes：
(1) the micro- knowledge of search engine retrieving：Search engine is inputted using the title of micro- knowledge as keyword, from Baidupedia, dimension
Base encyclopaedia, Chinese encyclopaedia net obtain the knowledge entry related to micro- knowledge；
(2) integrated searching result：The retrieval result of acquisition is integrated, merges the content of repetition, each knowledge point is integrated
Into the brief content of 300 word or so；
(3) micro- knowledge is stored：For each micro- knowledge using search key as title, the content of integration is knowledge body, is stored in micro-
Knowledge base.
6. method is supported based on the theoretical personalized instant learning of constructive learning as described in claim 2, wherein structure expands
Exhibition knowledge base step includes：
(1) retrieval and micro- knowledge related web page：With the entitled search key of micro- knowledge, related web page is retrieved；
(2) screening retrieval content：Filtered out and micro- maximally related content of knowledge in the webpage obtained from retrieval；
(3) storage extension knowledge：With the title of the entitled extension knowledge of micro- knowledge, effective content of acquisition is retrieved as extension
Knowledge body, storage to extension knowledge base.
7. method is supported based on the theoretical personalized instant learning of constructive learning as described in claim 2, wherein finding to lack
The knowledge step of mistake includes：
(1) user knowledge background is obtained：The microblogging state of the newest issue of user is analyzed, grasps the knowledge point that user currently pays close attention to,
In this, as the knowledge background of user；
(2) user knowledge background is marked in label graph of a relation：By the current knowledge point of interest of user in label graph of a relation
It is marked；
(3) knowledge of missing is found：Existing relation between the knowledge point provided by label graph of a relation, find current with user
Knowledge point of interest other knowledge points in close relations, the as desired missing knowledge of user.
8. supporting method based on the theoretical personalized instant learning of constructive learning as described in claim 2, wherein knowledge pushes away
Recommending step includes：
(1) micro- knowledge to be recommended is determined：Using user it is desired missing knowledge first three as micro- knowledge to be recommended；
(2) extension knowledge to be recommended is determined：Fed back according to the study of the conventional login system of user, what which analysis user recommended to
Knowledge is interested, recommends corresponding extension knowledge for user；
(3) knowledge is recommended：Micro- knowledge to be recommended and extension knowledge recommendation to be recommended are learnt to user.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510252479.2A CN104866557B (en) | 2015-05-18 | 2015-05-18 | A kind of personalized instant learning theoretical based on constructive learning supports System and method for |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510252479.2A CN104866557B (en) | 2015-05-18 | 2015-05-18 | A kind of personalized instant learning theoretical based on constructive learning supports System and method for |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104866557A CN104866557A (en) | 2015-08-26 |
| CN104866557B true CN104866557B (en) | 2018-03-20 |
Family
ID=53912383
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510252479.2A Expired - Fee Related CN104866557B (en) | 2015-05-18 | 2015-05-18 | A kind of personalized instant learning theoretical based on constructive learning supports System and method for |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104866557B (en) |
Families Citing this family (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105989088B (en) | 2015-02-12 | 2019-05-14 | 马正方 | Learning device under digitized environment |
| CN106503172A (en) * | 2016-10-25 | 2017-03-15 | 天闻数媒科技（湖南）有限公司 | The method and apparatus that learning path recommended by knowledge based collection of illustrative plates |
| CN106651701A (en) * | 2016-12-29 | 2017-05-10 | 山东科技大学 | Learning resource constructing method and device |
| CN109947808A (en) * | 2017-08-15 | 2019-06-28 | 上海颐为网络科技有限公司 | A kind of intelligent learning system and method |
| CN107741978A (en) * | 2017-10-13 | 2018-02-27 | 北京中教在线科技有限公司 | A kind of Pushing personality study resource method and its system |
| CN107808014B (en) * | 2017-11-06 | 2020-02-21 | 北京中科智营科技发展有限公司 | Knowledge base establishing method based on natural language processing |
| TWI679600B (en) * | 2018-02-05 | 2019-12-11 | 多利曼股份有限公司 | System and method for characteristics prediction |
| CN111353091B (en) | 2018-12-24 | 2024-10-11 | 北京三星通信技术研究有限公司 | Information processing method, information processing device, electronic equipment and readable storage medium |
| CN109874032B (en) * | 2019-03-07 | 2021-06-22 | 四川长虹电器股份有限公司 | Program topic personalized recommendation system and method for smart television |
| CN110442670B (en) * | 2019-06-11 | 2023-05-26 | 天津交通职业学院 | Consumer portrait generation method based on text indexing |
| CN117573985B (en) * | 2024-01-16 | 2024-04-05 | 四川航天职业技术学院(四川航天高级技工学校) | Information pushing method and system applied to intelligent online education system |
| CN118606563B (en) * | 2024-08-08 | 2024-10-08 | 光谷技术有限公司 | Personalized interaction method and system based on database platform |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1670727A (en) * | 2004-08-12 | 2005-09-21 | 金德龙 | Knowledge intension based knowledge information retrieval method and system thereof |
| TW200807346A (en) * | 2006-07-17 | 2008-02-01 | Hamastar Technology Co Ltd | Knowledge framework system and method for integrating a knowledge management system with an e-learning system |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20030195834A1 (en) * | 2002-04-10 | 2003-10-16 | Hillis W. Daniel | Automated online purchasing system |
| US20120329030A1 (en) * | 2010-01-29 | 2012-12-27 | Dan Joseph Leininger | System and method of knowledge assessment |
-
2015
- 2015-05-18 CN CN201510252479.2A patent/CN104866557B/en not_active Expired - Fee Related
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1670727A (en) * | 2004-08-12 | 2005-09-21 | 金德龙 | Knowledge intension based knowledge information retrieval method and system thereof |
| TW200807346A (en) * | 2006-07-17 | 2008-02-01 | Hamastar Technology Co Ltd | Knowledge framework system and method for integrating a knowledge management system with an e-learning system |
Non-Patent Citations (1)
| Title |
|---|
| 基于标签协同过滤算法在微博推荐中的研究;胡大伟;《中国优秀硕士学位论文全文数据库 信息科技辑》;20130515(第05期);I138-2120 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104866557A (en) | 2015-08-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104866557B (en) | A kind of personalized instant learning theoretical based on constructive learning supports System and method for | |
| CN104933113B (en) | A kind of expression input method and device based on semantic understanding | |
| US10650188B2 (en) | Constructing a narrative based on a collection of images | |
| CN110633373B (en) | Automobile public opinion analysis method based on knowledge graph and deep learning | |
| CN103425799B (en) | Individuation research direction commending system and recommend method based on theme | |
| CN103324665B (en) | Hot spot information extraction method and device based on micro-blog | |
| JP4637969B1 (en) | Properly understand the intent of web pages and user preferences, and recommend the best information in real time | |
| CN107578292B (en) | User portrait construction system | |
| Foley et al. | Learning to extract local events from the web | |
| CN104484431B (en) | A kind of multi-source Personalize News webpage recommending method based on domain body | |
| CN108804701A (en) | Personage's portrait model building method based on social networks big data | |
| JP2011248831A (en) | Information processor and information processing method, and program | |
| CN110134792A (en) | Text recognition method, device, electronic equipment and storage medium | |
| CN113407729B (en) | Judicial-oriented personalized case recommendation method and system | |
| Chowdury et al. | A data mining based spam detection system for youtube | |
| CN107203520A (en) | The method for building up of hotel's sentiment dictionary, the sentiment analysis method and system of comment | |
| CN111723256A (en) | Government affair user portrait construction method and system based on information resource library | |
| CN110110218B (en) | Identity association method and terminal | |
| CN112036659A (en) | Social network media information popularity prediction method based on combination strategy | |
| CN109684548A (en) | A kind of data recommendation method based on user's map | |
| US9544384B2 (en) | Method and system for pushing associated users in social networking service network | |
| CN105608118B (en) | Result method for pushing based on customer interaction information | |
| CN106021251A (en) | Hierarchical semantic model image retrieval method based on background knowledge | |
| Marchi et al. | Measuring destination image of an Italian island: An analysis of online content generated by local operators and tourists | |
| CN110175289A (en) | Mixed recommendation method based on cosine similarity collaborative filtering |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| EXSB | Decision made by sipo to initiate substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20210723 Address after: 211500 floor 6, building a, hatch Eagle building, No. 99, Tuanjie Road, Jiangbei new area, Nanjing, Jiangsu Patentee after: Jiangsu Guoji Technology Co.,Ltd. Address before: No. 1800 road 214122 Jiangsu Lihu Binhu District City of Wuxi Province Patentee before: Jiangnan University |
|
| 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: 20180320 |