CN104598474B - Information recommendation method based on data semantic under cloud environment - Google Patents
Information recommendation method based on data semantic under cloud environment Download PDFInfo
- Publication number
- CN104598474B CN104598474B CN201310532501.XA CN201310532501A CN104598474B CN 104598474 B CN104598474 B CN 104598474B CN 201310532501 A CN201310532501 A CN 201310532501A CN 104598474 B CN104598474 B CN 104598474B
- Authority
- CN
- China
- Prior art keywords
- information
- semantic
- user preference
- basic data
- preference information
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
- Machine Translation (AREA)
Abstract
The present invention relates to the information recommendation method based on data semantic under a kind of cloud environment, this method realizes information recommendation by magnanimity semantic information index module under the semantization module of basic data and user preference information, cloud environment and the information recommendation module based on semantic computation, the wherein semantization module of basic data and user preference information, basic data and user preference information are obtained by cloud platform, and semantization description is carried out to basic data and user preference information, build the ontology library of basic data and user preference information；Magnanimity semantic information index module under cloud environment, to the information structuring index structure of semantization, and when index node overloads, the division and restructuring that are indexed；Information recommendation module based on semantic computation, carries out semantic computation to the body of basic data and user preference information, obtains information recommendation result.Compared with prior art, the present invention has the advantages that real-time is good, robustness is high, recommends quality high.
Description
Technical field
The present invention relates to the data processing method under a kind of cloud environment, more particularly, to being based on data language under a kind of cloud environment
The information recommendation method of justice.
Background technology
Internet and technology of Internet of things are developed rapidly so that magnanimity information is presented in face of us at the same time, for example, when working as
There are nearly million books, Netflix to there are millions of films, eBay nets to have millions of new publication things daily on line on the net
Product, and 1,500,000,000 web page storage is had more than above del.icio.us community networks, information overload is in outburst trend, its result is led
Terminal user has been caused can not accurately and efficiently to find oneself object interested.Therefore, for enterprise, information overload is asked
Topic will seriously reduce the economic benefit and the market competitiveness of its own.At present, information recommendation system is to solve information overload problem
One of most effective instrument.Under the competitive environment being growing more intense, information recommendation system has been not only a kind of battalion of business
Pin means, it is often more important that the tackness of user can be promoted
In recent years, researcher is placed on the research center of gravity of information recommendation system in the design of recommendation method, this is mainly
Because recommending the core that method is information recommendation system, it decides the quality of system performance.At present, the recommendation side of mainstream
Method has 3 classes, i.e. content-based recommendation method, collaborative filtering recommending method and mixing recommendation method.
Content-based recommendation method (Content-based Recommendation) is mainly using artificial intelligence, data
The technology such as excavation and probability statistics, defines object, simultaneity factor is based on user's evaluation object by the attribute of correlated characteristic
Feature learns the interest of user, and the matching degree according to subscriber data and object to be predicted is recommended, make great efforts to push away to client
Recommend the product similar to the product liked before it.At present, this respect has SKCBR methods, ADL side than more typical recommendation method
Method, SNP methods and YSNLG methods.Greatest drawback based on content recommendation method is：It must analysis product content information,
Therefore it is helpless to contents such as music, image, videos, the quality of its information can not be analyzed.Therefore, this kind of method should in reality
There are significant limitation in.
Collaborative filtering recommending method (Collaborative Filtering Recommendation) tracks and using use
The historical information at family calculates the similitude between user, and then, other are produced using neighbours higher with targeted customer's similitude
Product are evaluated to predict fancy grade of the targeted customer to specific products, and finally, system is according to this fancy grade come to target
User is recommended.At present, collaborative filtering recommending method has two kinds of different types：Side based on memory (Memory-based)
The method of method and (Model-based) based on model.Undue product is beaten according to all in system based on the method for memory
Information is predicted, and learn simultaneously recommended user's behavior mould using the marking data being collected into based on the method for model
Type, and then marking is predicted to some product.The defects of collaborative filtering recommending method is maximum be：With product and number of users
Increase, the time complexity of method will exponentially increase, so as to cause system can not in real time or quickly recommend suitable production
Product are to user.In addition, this kind of method is poor for the recommendation effect of the user of new registration and the Xin product reached the standard grade.
The defects of in order to make up content-based recommendation method and collaborative filtering recommending method each, researcher proposes
Mix recommendation method (Hybrid Recommendation).It is main for different application scenarios and demand, mixing recommendation method
It is divided into 3 classes：(1) integrate afterwards：Recommendation list is respectively obtained using content-based recommendation method and collaborative filtering recommending method, is collected
Result into list determines the object finally recommended.Claypool M et al. have used the linear combination of appraisal result, and
Pazzani M et al.] voting mechanism has been used to combine these recommendation results.(2) integrated in：In a kind of recommendation method as frame
Frame, integrates another kind recommendation method.Soboroff I et al. use LSI (Latent Semantic Indexing) Indexing Mechanism
The user characteristics vector refined is used in the frame based on content.And Good N et al. user as object, by based on
The feature extracting method of content uses the feature of user in itself in similarity measure, rather than only relies only on the click of user
Behavior.(3) integrated before：Directly the method based on content and collaborative filtering is incorporated under a unified frame model.Basu
The age of user and the type of film are put into training study in a unified grader by C et al..And Ansari A et al. make
The parameter that this mould ploughs is obtained with Bayes's mixed effect regression model, and by markov monte carlo method.
However, it has been found that the maturation of 2.0 technology of appearance and Web with mass data, existing information recommendation
Technology at least faces three serious problems：(1) information of substantial amounts of user and product is that dynamic changes on website, which results in
Existing recommendation method needs costly calculation amount to model again, so as to seriously affect the real-time of recommendation results；(2)
2.0 network openings of Web cause website to be often subject to the attack of malicious user, and customer flow pressure causes software module
Exception, and the robustness of existing information recommendation method is weaker, this causes system to be easy to paralyse；(3) existing information recommendation
Method is usually only modeled the current preference of user, and without concern for the evolution process of preference, this is largely by shadow
Ring the quality recommended and personalized adaptive effect.
The content of the invention
It is an object of the present invention to overcome the above-mentioned drawbacks of the prior art and provide be based under a kind of cloud environment
The information recommendation method of data semantic.
The purpose of the present invention can be achieved through the following technical solutions：
Information recommendation method based on data semantic under a kind of cloud environment, this method are believed by basic data and user preference
Magnanimity semantic information index module and the information recommendation module based on semantic computation are realized under the semantization module of breath, cloud environment
Information recommendation, wherein
The basic data and the semantization module of user preference information, basic data and user are obtained by cloud platform
Preference information, and semantization description is carried out to basic data and user preference information, build basic data and user preference information
Ontology library；
Magnanimity semantic information index module under the cloud environment, to the information structuring index structure of semantization, and in rope
When drawing node overload, the division and restructuring that are indexed；
The information recommendation module based on semantic computation, language is carried out to the body of basic data and user preference information
Justice calculates, and obtains information recommendation result.
The basic data and the semantization module of user preference information, formalization table is carried out to the semanteme of basic data
Show, and thus build basic data ontology library；Dominant semantics extraction and Latent Semantic are carried out to user preference information at the same time to find,
And user preference information ontology library is built by dominant semantic and Latent Semantic.
The body of basic data represents that wherein C represents concept in basic data using five-tuple O=(C, R, P, I, A)
The set of term, R are the polynary mapping on C × C to R, i.e. set of relationship between concept, and P is the category for illustrating concept characteristic
Property set, I is the example collection of concept, and A is regular collection.
The dominant semantics extraction of user preference information selects text using potential applications index and support vector machines technology
Concept in washer section, so as to complete dominant semantics extraction；The dominant semantic discovery of user preference information is then from the semanteme of selection
Concept is set out, other relevant concepts of the semantic concept for going out and choosing using the basic data ontological analysis produced,
Relation, attribute and example, complete stealthy semantic discovery.
The index structure is the two-stage distributed index structure based on CAN and CHORD hybrid routing protocols, wherein entirely
Office's index is distributed on several servers in cloud platform, the global index's segment safeguarded simultaneously for each server, according to
Corresponding partial indexes are stored according to the particular server cluster specified based on CAN and CHORD hybrid routing protocols.
The information recommendation module based on semantic computation carries out information recommendation using two methods：
1) using basic data body and user preference information body as input, with basic data body and user preference information
Body carries out the semantic computation based on algebra on ontology, retains the basic data sheet that user preference information body similarity is higher than threshold value
Body, carries out information recommendation；
2) first by the user's preference information body and the association user preference information body obtained from community network into
Row semantic computation, retains the association user preference information body for being higher than threshold value with user preference information body similarity, as pass
Join preference body；Obtained association preference body and basic data body are subjected to the semantic computation based on algebra on ontology again, protected
The basic data body for being higher than threshold value with associating preference body similarity is stayed, carries out information recommendation.
The information recommendation module based on semantic computation further includes user preference evolution chain, the user's preference evolution chain
It is made of the user preference information body of different time node, the preference situation of change of recording and tracking different times user.
Compared with prior art, the present invention has the following advantages：
1st, when dynamic, which occurs, in the information of the user on website and product to be changed, it is not necessary to expend huge calculation amount to weigh
New modeling, so as to improve the real-time of recommendation results.
2nd, the information recommendation system robustness based on this method is stronger, is not easy to paralyse.
3rd, the present invention includes user preference evolution chain, it is possible to increase the quality of information recommendation and personalized adaptive effect.
Brief description of the drawings
Fig. 1 is the schematic diagram of the present invention.
Embodiment
The present invention is described in detail with specific embodiment below in conjunction with the accompanying drawings.
Embodiment
As shown in Figure 1, the information recommendation method based on data semantic under a kind of cloud environment, this method by basic data and
Magnanimity semantic information index module and the information based on semantic computation push away under the semantization module of user preference information, cloud environment
Recommend module and realize information recommendation, wherein
The semantization module of basic data and user preference information obtains basic data by cloud platform and user preference is believed
Breath, and semantization description is carried out to basic data and user preference information, build the body of basic data and user preference information
Storehouse.
Basic data body represents that wherein C represents concept term in basic data with five-tuple O=(C, R, P, I, A)
Set；R is the polynary mapping on C × C to R, i.e. set of relationship between concept；P is the property set for illustrating concept characteristic
Close；I is the example collection of concept；A is regular collection.Expression model based on basic data body, in order to improve it is intelligent and
Knowledge reuse rate, the present invention devise the virtual domain body constructing technology of parallelization, which passes through to initial storage in system
Most basic basic data body carry out stratification processing, document is described using the field of DODL language, based on biotic population
They are combined or deleted by evolutionary mechanism, dynamically construct global body.
Secondly, the present invention obtains name Entity recognition (the Named Entity in field using biological knowledge
Recognition) mechanism, devises two-stage model and carrys out semantization user preference information.In the first stage of semantization, this hair
It is bright to regard specific user preference information as a document snippet, index (LSI using potential applications:Latent Semantic
) and support vector machines (SVM Index:Support Vector Machine) technology come effectively select it is general in document snippet
Read, that is, complete dominant semantic extraction.And in second stage, the present invention is from the concept chosen, using having produced
Basic data ontological analysis relevant other concepts, relation, attribute and the example of the concept that go out and choose, that is, complete hidden
Fractal semantic is found.On the basis of dominant semantic and Latent Semantic, the sheet of the user's preference information is automatically or semi-automatically constructed
Body.
Magnanimity semantic information index module is to the information structuring index structure of semantization under cloud environment, and in index node mistake
During load, the division and restructuring that are indexed.
In terms of the implementation of index framework, the present invention constructs the two-stage based on C2 (CAN and CHORD) hybrid routing protocol
Distributed index structure C 2-DISINX.In the index structure, global index is distributed in several clothes in cloud platform by we
It is engaged on device, and to global index's segment that each server is safeguarded, specific server is specified according to C2 hybrid routing protocols
Cluster stores corresponding partial indexes.In this way, it is possible to fundamentally ensure the scalability and height of distributed index
Effect property.
When magnanimity semantization information updates, incremental maintenance C2-DISINX index structures of the present invention.According to distribution
Data base theory understands that the groundwork of index maintenance is to handle the division and merging of C2-DISINX index nodes.The present invention carries
Go out a near-optimization strategy to select to need the index node for dividing and merging, its embodiment is (it is assumed that storage is complete
The node set of office's index and partial indexes is respectively W and S)：Construction weights oriented bigraph (bipartite graph) first, by index node collection W and S
The vertex set of bigraph (bipartite graph) is mapped as, and the cost information being route between the node obtained by sampling assessment is mapped as the side of bigraph (bipartite graph)
Collection.Figure shortest path first theory is next based on, by introducing a virtual vertex, in constant time complexity, will be weighted
Oriented bigraph (bipartite graph) is converted to Steiner weight path figures.Finally, generated in polynomial time complexity oriented in path profile
Steiner trees, and then obtain the approximate optimal solution for the index node for needing to divide and merging.According to oriented Steiner tree theories,
The time complexity PT of near-optimization strategy can be adjusted and weighed by the positive number not less than 1 with optimization lower bound OB.
Information recommendation module based on semantic computation carries out semantic computation to the body of basic data and user preference information,
Obtain information recommendation result.
In terms of algebra on ontology semantic computation is implemented, the present invention proposes EOAS algebra on ontology systems (Extension
Ontology Algebra System), which is defined as：∑=(O, R, Op '), wherein, ∑ is intrinsic generation
Number, O are Ontological concept set, and R is four kinds of relations (part-of, kind-of, attribute-of and instance-of), Op '
In addition to comprising original intersecting and merging, poor three kinds of set operation operators, the arithmetical operation such as we add, subtract, multiplication and division is calculated
Son, with or, the logical operation operator such as non-.In the algebra system, the arithmetical operation operator added is to be directed to body similarity
Arithmetical operation needed for calculating process is set, and logical operation operator is patrolled for needed for Similarity measures process
Collect and judge computing to set.It is of the invention on the basis of the semantic description of sequential logic in terms of semantic temporal calculating is implemented, if
Count a set of sequential and calculated operator, realize parallel, order semantically, interrupt, recover, hanging up etc. that operation calculates.Meanwhile we
The rule such as calculation law of temporal operator will be also defined, makes in information recommendation system different objects when carrying out by these rules
Correctness is kept during sequence semantic computation.
Based on algebra on ontology semantic computation, the present invention is inputted using basic data body and user preference body as algorithm,
Under the support of EOAS algebra on ontology systems, the semantic meter based on algebra on ontology is carried out to basic data body and user preference body
Calculate, high object is retained with user preference body similarity, and the low object of similarity is then filtered, so that effectively
Implementation information is recommended.Especially, in order to effective integration society recommendation mechanisms, the present invention first by the user's preference body and from
The preference body that associated user is analyzed in community network carries out semantic computation, then, then the association preference body that will be obtained
The semantic computation based on algebra on ontology is carried out with basic data body, to draw final recommendation results.In addition, in order to more effective
Personalized adaptive recommendation is carried out, the present invention proposes user preference evolution chain technology.User preference evolution chain is by different time
The user preference body composition of node, the preference situation of change of recording and tracking different times user, and then by analyzing and digging
User's object interested is more accurately predicted and adjusted to knowledge in pick preference evolution chain.Drilled in designing user preference body
Change in link, the body interior element that preference Ontology Evolution is included including concept, relation, attribute and rule by the present invention is met
To inconsistent judgement, deletion, addition, the evolution process equivalence such as succession be changed into it is parallel, suitable in sequential logic semantic computation
Sequence, interruption, recovery and hang-up etc. operate.
Claims (6)
- A kind of 1. information recommendation method based on data semantic under cloud environment, it is characterised in that this method by basic data and Magnanimity semantic information index module and the information based on semantic computation push away under the semantization module of user preference information, cloud environment Recommend module and realize information recommendation, whereinThe basic data and the semantization module of user preference information, basic data and user preference are obtained by cloud platform Information, and semantization description is carried out to basic data and user preference information, build the sheet of basic data and user preference information Body storehouse；Magnanimity semantic information index module under the cloud environment, is saved to the information structuring index structure of semantization, and in index During point overload, the division and restructuring that are indexed；The information recommendation module based on semantic computation, semantic meter is carried out to the body of basic data and user preference information Calculate, obtain information recommendation as a result, the information recommendation module based on semantic computation includes user preference evolution chain, the user Preference evolution chain is made of the user preference information body of different time node, and the preference of recording and tracking different times user becomes Change situation.
- 2. the information recommendation method based on data semantic under a kind of cloud environment according to claim 1, it is characterised in that institute The basic data and the semantization module of user preference information stated, formalization representation is carried out to the semantic of basic data, and thus Build basic data ontology library；Dominant semantics extraction and Latent Semantic are carried out to user preference information at the same time to find, and by dominant Semantic and Latent Semantic structure user preference information ontology library.
- 3. the information recommendation method based on data semantic under a kind of cloud environment according to claim 2, it is characterised in that base The body of plinth data represents that wherein C represents the set of concept term in basic data using five-tuple O=(C, R, P, I, A), R is the polynary mapping on C × C to R, i.e. set of relationship between concept, and P is the attribute set for illustrating concept characteristic, and I is general The example collection of thought, A are regular collections.
- 4. the information recommendation method based on data semantic under a kind of cloud environment according to claim 2, it is characterised in that use The dominant semantics extraction of family preference information is selected in document snippet using potential applications index and support vector machines technology Concept, so as to complete dominant semantics extraction；The dominant semantic discovery of user preference information is then from the semantic concept of selection, profit Other relevant concepts of semantic concept for being gone out and chosen with the basic data ontological analysis produced, relation, attribute and Example, completes stealthy semantic discovery.
- 5. the information recommendation method based on data semantic under a kind of cloud environment according to claim 1, it is characterised in that institute The index structure stated is the two-stage distributed index structure based on CAN and CHORD hybrid routing protocols, and wherein global index is distributed On several servers in cloud platform, global index's segment for being safeguarded simultaneously for each server, according to based on CAN and Particular server cluster that CHORD hybrid routing protocols are specified stores corresponding partial indexes.
- 6. the information recommendation method based on data semantic under a kind of cloud environment according to claim 1, it is characterised in that institute The information recommendation module based on semantic computation stated carries out information recommendation using two methods：1) using basic data body and user preference information body as input, with basic data body and user preference information body The semantic computation based on algebra on ontology is carried out, retains the basic data body that user preference information body similarity is higher than threshold value, Carry out information recommendation；2) the user's preference information body and the association user preference information body obtained from community network are subjected to language first Justice calculates, and retains the association user preference information body for being higher than threshold value with user preference information body similarity, inclined as association Good body；Obtained association preference body and basic data body are subjected to the semantic computation based on algebra on ontology again, retain with The basic data body that preference body similarity is higher than threshold value is associated, carries out information recommendation.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310532501.XA CN104598474B (en) | 2013-10-30 | 2013-10-30 | Information recommendation method based on data semantic under cloud environment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310532501.XA CN104598474B (en) | 2013-10-30 | 2013-10-30 | Information recommendation method based on data semantic under cloud environment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104598474A CN104598474A (en) | 2015-05-06 |
| CN104598474B true CN104598474B (en) | 2018-04-27 |
Family
ID=53124267
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310532501.XA Expired - Fee Related CN104598474B (en) | 2013-10-30 | 2013-10-30 | Information recommendation method based on data semantic under cloud environment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104598474B (en) |
Families Citing this family (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104850632A (en) * | 2015-05-22 | 2015-08-19 | 东北师范大学 | Generic similarity calculation method and system based on heterogeneous information network |
| US10146873B2 (en) * | 2015-06-29 | 2018-12-04 | Microsoft Technology Licensing, Llc | Cloud-native documents integrated with legacy tools |
| CN105023178B (en) * | 2015-08-12 | 2018-08-03 | 电子科技大学 | A kind of electronic commerce recommending method based on ontology |
| CN107066582B (en) * | 2017-04-14 | 2020-06-26 | 聚好看科技股份有限公司 | Method and device for realizing virtual resource recommendation |
| CN108304523B (en) * | 2017-10-23 | 2021-11-09 | 同济大学 | Method for extracting and standardizing text time constructed for knowledge graph |
| CN108664943B (en) * | 2018-05-17 | 2020-06-05 | 北京科技大学 | Indoor daily activity recognition method under multi-occupant scene |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1632792A (en) * | 2004-12-29 | 2005-06-29 | 复旦大学 | XML data based highly effective path indexing method |
| CN101770520A (en) * | 2010-03-05 | 2010-07-07 | 南京邮电大学 | User interest modeling method based on user browsing behavior |
| CN102663010A (en) * | 2012-03-20 | 2012-09-12 | 复旦大学 | Personalized image browsing and recommending method based on labelling semantics and system thereof |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20120052636A (en) * | 2010-11-16 | 2012-05-24 | 한국전자통신연구원 | A hscode recommendation service system and method using ontology |
| US8620964B2 (en) * | 2011-11-21 | 2013-12-31 | Motorola Mobility Llc | Ontology construction |
-
2013
- 2013-10-30 CN CN201310532501.XA patent/CN104598474B/en not_active Expired - Fee Related
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1632792A (en) * | 2004-12-29 | 2005-06-29 | 复旦大学 | XML data based highly effective path indexing method |
| CN101770520A (en) * | 2010-03-05 | 2010-07-07 | 南京邮电大学 | User interest modeling method based on user browsing behavior |
| CN102663010A (en) * | 2012-03-20 | 2012-09-12 | 复旦大学 | Personalized image browsing and recommending method based on labelling semantics and system thereof |
Non-Patent Citations (1)
| Title |
|---|
| "Indexing multi-dimensional data in a cloud system";Jinbao Wang等;《Proceeding SIGMOD "10 Proceedings of the 2010 ACM SIGMOD International Conference on Management of data》;20100610;第591-602页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104598474A (en) | 2015-05-06 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104598474B (en) | Information recommendation method based on data semantic under cloud environment | |
| Ristoski et al. | Mining the web of linked data with rapidminer | |
| TWI496015B (en) | Text matching method and device | |
| US11170306B2 (en) | Rich entities for knowledge bases | |
| Dubey et al. | Item-based collaborative filtering using sentiment analysis of user reviews | |
| CN111259220B (en) | Data acquisition method and system based on big data | |
| Tibély et al. | Extracting tag hierarchies | |
| US20140114902A1 (en) | Intelligent Data Agent for a Knowledge Management System | |
| US20160364379A1 (en) | Web-based influence system and method | |
| Kanavos et al. | Graph Community Discovery Algorithms in Neo4j with a Regularization-based Evaluation Metric. | |
| US20140114949A1 (en) | Knowledge Management System | |
| US9305261B2 (en) | Knowledge management engine for a knowledge management system | |
| US9720984B2 (en) | Visualization engine for a knowledge management system | |
| CN108154380A (en) | The method for carrying out the online real-time recommendation of commodity to user based on extensive score data | |
| Jia et al. | A privacy-sensitive service selection method based on artificial fish swarm algorithm in the internet of things | |
| Lasso et al. | Towards an alert system for coffee diseases and pests in a smart farming approach based on semi-supervised learning and graph similarity | |
| Mohana et al. | Optimized service discovery using qos based ranking: A fuzzy clustering and particle swarm optimization approach | |
| CN113569162A (en) | Data processing method, device, equipment and storage medium | |
| Zhang et al. | Risk management of commodity trade business based on deep learning and parallel processing of visual multimedia big data | |
| Katsimpras et al. | Class-aware tensor factorization for multi-relational classification | |
| Wang et al. | Random partition factorization machines for context-aware recommendations | |
| Barbieri et al. | Survival factorization on diffusion networks | |
| Ashraf et al. | Making sense from Big RDF Data: OUSAF for measuring ontology usage | |
| Saouli et al. | SaaS-DCS: software-as-a-service discovery and composition system-based existence degree | |
| Luthfi et al. | Toward a Reference Architecture for User-Oriented Open Government Data Portals |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into 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 |
Granted publication date: 20180427 Termination date: 20201030 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |