CN102681979A - Content editing intelligent verifying method facing to open knowledge community - Google Patents
Content editing intelligent verifying method facing to open knowledge community Download PDFInfo
- Publication number
- CN102681979A CN102681979A CN2012101508089A CN201210150808A CN102681979A CN 102681979 A CN102681979 A CN 102681979A CN 2012101508089 A CN2012101508089 A CN 2012101508089A CN 201210150808 A CN201210150808 A CN 201210150808A CN 102681979 A CN102681979 A CN 102681979A
- Authority
- CN
- China
- Prior art keywords
- degree
- semantic
- resource
- belief
- trust
- 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.)
- Granted
Links
Images
Landscapes
- Machine Translation (AREA)
Abstract
The invention discloses a content editing intelligent verifying method facing to an open knowledge community. The content editing intelligent verifying method comprises the following steps of: (1), extracting the semantic feature information of a resource content and calculating a semantic similarity between the feature item aggregate and the semantic gene of a content newly added; (2), based on the interactive operation data of a user in the open knowledge community, calculating the trust degree of the user by applying a trust evaluation model; and (3), synthesizing two indexes of the semantic similarity and the trust degree of the user and judging whether to automatically accept or refuse content editing for this time. By using the content editing intelligent verifying method provided by the invention, the burden of manually verifying the content editing in the open knowledge community can be effectively lightened, and further, the content editing intelligent verifying method has a quite high accuracy rate.
Description
Technical field
The invention belongs to computer science and applied technical field, specifically, a kind of Edition Contains intelligence checking method towards open Knowledge Community is applied to the Resources Construction and the management of various open Knowledge Community.
Background technology
In recent years, be that the open Knowledge Community of representative becomes more and more popular with wikipedia (Wikipedia), external have Google Knol, Cloudworks, a Freebase etc., and Baidu is known in domestic having, Baidu's encyclopaedia, interactive encyclopaedia, study unit platform etc.Open knowledge platform no doubt has its special advantages, can attract that a large number of users is participated in, the performance group wisdom, promote knowledge transmission mechanism and share.Yet, also brought unavoidable trouble in the time of open.The q&r problem that editor that wikipedia is open fully and organizational form make resource in the open Knowledge Community becomes the focus of concern.The Web2.0 epoch everybody can produce, consumption, transmission resource, user group's the complicacy and the liberalization of production directly cause explosive growth and resource quality very different of resource quantity in the open Knowledge Community.
In order to guarantee the reliability of the resource quality in the open Knowledge Community, current various open Knowledge Community mainly adopt the manual decision and the mode of manually Edition Contains audit to realize the control to the resource content quality.What wikipedia adopted is the orderly evolution that a cover is realized information based on the coordination system and a series of constraint rule (3R rule, true verification and evaluation rule at the same level in real time etc.) of manual work cooperation.Other various open Knowledge Community (Baidu's encyclopaedia, Google Knol, Freebase etc.) are also continued to use the control model of wikipedia basically, on the basis of feedback, interchange, finally realize the continuous change and the evolution of contents version through the mode of manual examination and verification.In recent years, researcher (Javanmardi et al., 2010 are arranged; Moturu&Liu, 2009) begin to attempt making up the trust evaluation model in the wikipedia, help the user and screen high-quality resource.But these trust evaluation models are based on mostly that compiles historical data makes up, and have ignored other abundant more interactivity information that help judges and resource degree of belief, such as service datas such as collecting, subscribe to, share.
Along with the continuous expansion of resource colony and user group's scale, the workload of Edition Contains audit will increase rapidly, bring huge work load must for creator, the supvr of knowledge.Current manual decision and content auditing technology can't adapt to the needs of open Knowledge Community development, therefore, are badly in need of exploring and a kind ofly can effectively alleviate the Intelligentized method that user in the open Knowledge Community carries out Edition Contains audit burden and pressure.
Summary of the invention
The technical matters that the present invention will solve is: the deficiency that overcomes existing manual examination and verification Edition Contains; A kind of intelligentized Edition Contains checking method is provided; This method can effectively alleviate the burden of manual examination and verification Edition Contains in the open Knowledge Community, and has higher accuracy rate.
The technical solution adopted for the present invention to solve the technical problems is: a kind of Edition Contains intelligence checking method towards open Knowledge Community is characterized in that may further comprise the steps:
(1) the semantic feature information (semantic gene) of extraction resource content is calculated the new characteristic item of content and the semantic similarity of semantic gene of adding;
(2) based on user's interactive operation data, the application trust assessment models is calculated user's degree of belief;
(3) comprehensive semantic similarity and two indexs of users to trust degree judge whether to accept this Edition Contains.
Semantic gene in the above-mentioned steps (1) is meant the education resource inherent structure of knowledge behind, can reflect the core content that resource institute will express, can formalization representation for gather based on the notion that has weight of ontology describing and notion between semantic relation.Semantic gene method for distilling is: the characteristic item that extracts resource content based on domain body; Weight according to characteristic evaluating function calculation characteristic item; Characteristic item is mapped as the notion in the domain body; Extract the semantic relation that notion exists based on the JENA framework in ontology library.Described characteristic evaluating function adopts word frequency statistics TF (Term Frequency) method.
The characteristic item collection of the new interpolation content in the above-mentioned steps (1) and the semantic similarity computing method of semantic gene are: the synonym speech woods of using Harbin Institute of Technology's extended edition is replaced the synonym in characteristic item collection and the semantic gene; Adopt cosine coefficient method computing semantic similarity.
Trust evaluation model core design thinking in the above-mentioned steps (2) is: comprise two core components of resource degree of belief and users to trust degree; The resource degree of belief is according to directly degree of belief evaluating data (user directly gives a mark to the resource degree of belief) and indirect degree of belief evaluating data (user is to the interactive operation information of resource, like subscription, collection etc.) adopt weighted-average method to calculate; The users to trust degree adopts weighted-average method to calculate according to the confidence level of the interactive information between the user and its creation knowledge; Influence each other between resource degree of belief and the users to trust degree, adopt the two degree of belief of iterative approximation calculated crosswise.The iterative approximation core concept is through setting a maximum error value max_error; Through iterative computation repeatedly; To the absolute value of the difference of all corresponding in twice result of calculation in front and back degree of beliefs all less than max_error; Just finish computing, generate and approach real resource degree of belief and users to trust degree.
The determination methods whether Edition Contains in the above-mentioned steps (3) is accepted is: setting an Edition Contains can received lowest threshold AT; Semantic similarity and two indexs of users to trust degree are adopted the weighted sum method; If result of calculation more than or equal to AT, is then accepted this Edition Contains; Otherwise, refuse this Edition Contains.
The present invention's beneficial effect compared with prior art:
(1) the inventive method can realize the automatic audit of Edition Contains in the open Knowledge Community is alleviated the burden and the pressure of manual examination and verification greatly through semantic gene of integrated application and trust evaluation model, has higher accuracy simultaneously.
(2) the present invention is based on the semantic feature information (semantic gene) that domain body extracts resource content,, can reach better extraction effect than the text feature extraction technology of routine.
(3) the trust evaluation model of the present invention's proposition; Can effective evaluation open users to trust degree and resource degree of belief in the Knowledge Community, have computational data and enrich, consider the interaction relation between users to trust and resource trust, the characteristics such as trusting relationship in the approaching more real society.
Description of drawings
Fig. 1 is the leaching process of semantic gene of the present invention;
Fig. 2 is the characteristic item leaching process based on domain body of the present invention;
The trust evaluation model that Fig. 3 proposes for the present invention;
The iterative approximation that Fig. 4 proposes for the present invention solves the process of calculated crosswise problem;
Fig. 5 is an Edition Contains intelligence review process of the present invention.
Embodiment
Introduce the present invention in detail below in conjunction with accompanying drawing and embodiment.
A kind of Edition Contains intelligence checking method towards open Knowledge Community of the present invention, adopt following steps:
(1) extracts semantic gene, computing semantic similarity
The notion of semantic gene is meant the education resource inherent structure of knowledge behind, can reflect resource the core content that will express.Be different from the file characteristics item of text similarity in relatively, semantic gene is not simple keyword set, but the semantic concept network that resource is hidden behind.
Semantic gene can be expressed as orderly tlv triple formally, i.e. SG=< CS, WS, RS >, and wherein CS is the key concept set, set sizes is no more than 10, CS={C1, C2, C3 ..., Cn}, 1≤n≤10; WS is the weight set of notion item, WS={W1, and W2, W3 ..., Wn}, wherein Wi is the weight of Ci, 1≤i≤n, and all weight sums are 1; RS is the set of relations between key concept, RS={R1, R2, R3;, Rn}, each relation adopts RDF tlv triple < Subject, Predicate in the domain body; Object>expression, R1=< Concept1, Relationship, Concept2 >; Concept1 here and Concept2 not necessarily are included among the CS, can be other notions of field ontology library, and Relationship is the conceptual relation of from field ontology library, extracting.
The process that semantic gene extracts is as shown in Figure 1, comprises four key steps: the characteristic item based on domain body extracts; Weight according to characteristic evaluating function calculation characteristic item; The characteristic speech is to the mapping of Ontological concept; Extract the semantic relation that characteristic item (notion) exists based on increase income framework-JENA of semantic net in ontology library.Characteristic item extraction flow process based on domain body is seen Fig. 2, and the characteristic evaluating function adopts word frequency statistics TF (Term Frequency) method.
Algorithm 1 based on the characteristic speech of JENA to the mapping algorithm of Ontological concept (Term Mapping to Concept, TM2C)
Input: the characteristic item set TS={t of resource
i| i=1,2,3 ..., n}
Output: the notion set CS={C of resource
j| j=1,2,3 ..., n}
The false code of algorithm is following:
Algorithm 1 uses the popular JENA framework in Semantic Web field, and the characteristic item that extracts in the resource content is mapped as notion corresponding in the domain body one by one.
Algorithm 2 based on the conceptual relation of JENA extract (Concept Relationship Extraction, CRE)
Input: the notional word set CS={C of resource
j| j=1,2,3 ..., n}
Output: conceptual relation set RS={R
j| j=1,2,3 ..., m}
The false code of algorithm is following:
Algorithm 2 uses the popular JENA framework in Semantic Web fields, and the concept set that algorithm 1 is got access to is combined in the conceptual relation that exists in the domain body and extracts.
Algorithm 3 semantic gene extraction algorithms
Input: the Title of resource, Tag, Content and SemanticData
Output: the semantic gene SG=< CS, WS, RS>of resource
Committed step:
Step 1 is called ICTCLAS Title is carried out word segmentation processing and noise filtering
Step 2 is called ICTCLAS Tag is cut and noise filtering
Step 3 is called ICTCLAS Content is carried out the filtration of html label, word segmentation processing, noise filtering (removal function word)
Step 4 is obtained the body class in the semantic description information
The characteristic set of words that step 5 couple Step2 obtains in the Step5 is carried out the word combination in conjunction with domain body, discerns new characteristic speech
Step 6 is called inactive vocabulary, and the set of words that Step6 obtains is carried out the stop words filtration
Step 7 combines extended edition synonym speech woods of Harbin Institute of Technology and domain body to carry out the synonym replacement, obtains characteristic word set TS
Step 8 application characteristic evaluation function calculates the weight of each characteristic speech, obtains the weight set WS of characteristic speech
Step 9 is used the TM2C algorithm and is obtained notion set CS
Step 10 is used the CRE algorithm and is extracted conceptual relation set RS
Step 11 algorithm finishes, output CS, WS and RS
Adopt the cosine coefficient method to calculate the semantic gene of current education resource and the semantic similarity of the text feature item collection that the user newly adds content.The semantic gene of representing resource with X: X=(C
1, WC
1; C
2, WC
2; C
n, WC
n), C wherein
kBe the notion item in the semantic gene, WC
kBe C
kWeight, 1≤k≤n; The text feature vector of representing fresh content with Y: Y=(T
1, WT
1; T
1, WT
1; T
m, WT
m), T wherein
kBe the characteristic speech in the proper vector, WT
kBe T
kWeight, 1≤k≤m.Carry out before the calculating of cosine similarity, need the element among X and the Y be carried out the synonym replacement.The synonym replacement based on the synonym speech woods of Harbin Institute of Technology's extended edition built-in among the LCS, on the other hand, is searched the synonym among X and the Y in the conceptual relation tlv triple that can from the semantic gene of resource, comprise, and is replaced on the one hand.The semantic similarity computing formula is following:
(2) the application trust assessment models is calculated user's degree of belief
The present invention propose a kind of towards open Knowledge Community the trust evaluation model---(Twoway Interactive Feedback Model TIFM), sees Fig. 3 to the two-way interaction feedback model.GIFM comprises resource degree of belief and two core components of users to trust degree, and the two influences each other; Both sides are each item influence factors of degree of belief; The center is four hypothesis of relevant trust evaluation.Need to prove that the trust here refers to global trusting, but not the trusting relationship between two peer node in the P2P network.The resource degree of belief is represented the global trust evaluation of all community users to resource node, and the users to trust kilsyth basalt shows that every other user in the community is to active user's global trust evaluation.
The resource degree of belief (Resource Trust, RT) comprise direct degree of belief and indirect degree of belief two parts: directly (Direct Resource Trust DRT) calculates according to the dominance trust evaluation that the user carries out degree of belief; (Indirect Resource Trust, IRT) intersection record according to user and resource calculates in the degree of belief evaluation indirectly.At the asset creation initial stage, the number of times of participating in direct trust evaluation owing to the user is less, so DRT shared weight w in the total trust degree of resource is estimated is on the low side.The increasing function that weight w is is independent variable with direct trust evaluation number of times will be according to the change dynamics adjustment of direct evaluation number of times, and along with increasing of direct trust evaluation number of times, DRT will more and more can represent RT, and the w value also will improve gradually.The computing formula of resource degree of belief can be expressed as: RT=w * DRT+ (1-w) * IRT.DRT adopts method of weighted mean; IRT adopts the weighted sum method earlier, carries out normalization then and handles.
Users to trust degree in the open Knowledge Community is characterized by four-tuple: UT={UT
Res, UT
Col, UT
Fri, UT
Rev, UT
ResThe users to trust degree that expression is calculated by the degree of belief of user institute establishing resource, UT
ColThe users to trust degree that expression is calculated by cooperation relation between the user, UT
Fri, the users to trust degree that expression is calculated by the relation of the good friend between the user, UT
RevThe users to trust degree that expression is calculated by edit history.The computing formula of users to trust degree adopts the weighted sum method, can be expressed as: UT=UW1 * UT
Res+ UW2 * UT
Col+ UW3 * UT
Fri+ UW4 * UT
Rev
There is the problem of calculated crosswise in TIFM, i.e. the users to trust degree has been used in the calculating of resource degree of belief, and the degree of belief of resource has been used in the calculating of users to trust degree.The present invention adopts the iterative approximation head it off; Flow process is seen Fig. 4; The core thinking is the degree of belief through all resources in the iterative computation system repeatedly and all users; The absolute value of the difference of each item trust value all less than the maximum error value of setting, shows that the result of calculation of all trust values tends towards stability, near actual value in twice result of calculation in front and back.
(3) Edition Contains intelligence audit
The flow process of Edition Contains intelligence audit sees 5.The confidence level of Edition Contains is carried out weighted sum through semantic similarity and two indexs of users to trust degree and drawn, and formula can be expressed as: CT=W1 * SS+W2 * UT, SS represent semantic gene of resource and the new semantic similarity that adds content text characteristic item collection; UT representes user's degree of belief; W1 is the shared weight of semantic similarity, and W2 is the shared weight of users to trust degree, wherein W1+W2=1; CT ∈ [0,1].If CT is more than or equal to AT, then this time Edition Contains will be passed through by the automatic audit of system; Otherwise, will be refused automatically by system.
Wherein but AT is preset Edition Contains acceptance threshold.AT can adjust according to practical application effect.AT can be used for regulating and control the severe degree of Edition Contains intelligence audit, and AT is high more, and the Edition Contains that the user carries out is not easy by automatic acceptance more; AT is low more, and the Edition Contains that the user carries out will be not easy by automatic refusal more.
Claims (8)
1. Edition Contains intelligence checking method towards open Knowledge Community is characterized in that may further comprise the steps:
(1) the semantic feature information of extraction resource content is calculated the new characteristic item of content and the semantic similarity of semantic gene of adding;
(2) based on user's interactive operation data, the application trust assessment models is calculated user's degree of belief;
(3) comprehensive semantic similarity and two indexs of users to trust degree judge whether to accept this Edition Contains.
2. a kind of Edition Contains intelligence checking method according to claim 1 towards open Knowledge Community; It is characterized in that: the semantic gene in said step (1) is meant the education resource inherent structure of knowledge behind; Can reflect resource the core content that will express, formalization representation is for based on the notion set that has weight of ontology describing and the semantic relation between notion.
3. a kind of Edition Contains intelligence checking method according to claim 1 towards open Knowledge Community, it is characterized in that: the semantic gene method for distilling in said step (1) is: the characteristic item that extracts resource content based on domain body; Weight according to characteristic evaluating function calculation characteristic item; Characteristic item is mapped as the notion in the domain body; Extract the semantic relation that notion exists based on increase income framework-JENA framework of semantic net in ontology library.
4. a kind of Edition Contains intelligence checking method according to claim 1 towards open Knowledge Community, it is characterized in that: the characteristic item collection of the new interpolation content in said step (1) and the semantic similarity computing method of semantic gene are: use synonym speech woods the synonym in characteristic item collection and the semantic gene is replaced; Adopt cosine coefficient method computing semantic similarity.
5. a kind of Edition Contains intelligence checking method according to claim 1 towards open Knowledge Community, it is characterized in that: the trust evaluation model in said step (2) comprises resource degree of belief and two core components of users to trust degree; The resource degree of belief calculates according to the method for directly degree of belief evaluating data and degree of belief evaluating data employing weighted sum indirectly; Said direct degree of belief evaluating data is that the user directly gives a mark to the resource degree of belief, and said indirect degree of belief evaluating data is the interactive operation information of user to resource; The users to trust degree adopts the method for weighted sum to calculate according to the confidence level of the interactive information between the user and its creation knowledge; Influence each other between resource degree of belief and the users to trust degree, adopt the two degree of belief of iterative approximation calculated crosswise.
6. a kind of Edition Contains intelligence checking method according to claim 5 towards open Knowledge Community; It is characterized in that: described iterative approximation is through setting a maximum error value max_error; Through iterative computation repeatedly; To the absolute value of the difference of all corresponding in twice result of calculation in front and back degree of beliefs all less than max_error, just finish computing, generation approaches real resource degree of belief and users to trust degree.
7. a kind of Edition Contains intelligence checking method according to claim 1 towards open Knowledge Community, it is characterized in that: the method that judging whether in said step (3) accepted this Edition Contains is: setting an Edition Contains can received lowest threshold AT; Semantic similarity and two indexs of users to trust degree are adopted the weighted sum method; If result of calculation more than or equal to AT, is then accepted this Edition Contains; Otherwise, refuse this Edition Contains.
8. a kind of Edition Contains intelligence checking method towards open Knowledge Community according to claim 4 is characterized in that: the synonym speech woods that said synonym speech woods is Harbin Institute of Technology's extended edition.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210150808.9A CN102681979B (en) | 2012-05-15 | 2012-05-15 | Content editing intelligent verifying method facing to open knowledge community |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210150808.9A CN102681979B (en) | 2012-05-15 | 2012-05-15 | Content editing intelligent verifying method facing to open knowledge community |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102681979A true CN102681979A (en) | 2012-09-19 |
| CN102681979B CN102681979B (en) | 2015-04-22 |
Family
ID=46813932
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201210150808.9A Active CN102681979B (en) | 2012-05-15 | 2012-05-15 | Content editing intelligent verifying method facing to open knowledge community |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN102681979B (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102929931A (en) * | 2012-09-24 | 2013-02-13 | 上海师范大学 | Information credibility assessment method based on body in semantic net |
| CN109190342A (en) * | 2018-08-20 | 2019-01-11 | 济南大学 | The owner's auth method and community server of intelligence community |
| CN109254959A (en) * | 2018-08-17 | 2019-01-22 | 广东技术师范学院 | A kind of data evaluation method, apparatus, terminal device and readable storage medium storing program for executing |
| CN111815296A (en) * | 2020-07-23 | 2020-10-23 | 长沙公信诚丰信息技术服务有限公司 | Intelligent auditing method for third-party service system |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101446970A (en) * | 2008-12-15 | 2009-06-03 | 腾讯科技（深圳）有限公司 | Method for censoring and process text contents issued by user and device thereof |
| CN101923540A (en) * | 2010-07-20 | 2010-12-22 | 陈洁 | Language translation quality auditing method |
| CN102314453A (en) * | 2010-06-30 | 2012-01-11 | 百度在线网络技术（北京）有限公司 | Method and system for screening high-quality versions |
-
2012
- 2012-05-15 CN CN201210150808.9A patent/CN102681979B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101446970A (en) * | 2008-12-15 | 2009-06-03 | 腾讯科技（深圳）有限公司 | Method for censoring and process text contents issued by user and device thereof |
| CN102314453A (en) * | 2010-06-30 | 2012-01-11 | 百度在线网络技术（北京）有限公司 | Method and system for screening high-quality versions |
| CN101923540A (en) * | 2010-07-20 | 2010-12-22 | 陈洁 | Language translation quality auditing method |
Non-Patent Citations (3)
| Title |
|---|
| 李学明等: "P2P系统中基于双信任信息的信誉度改进模型", 《世界科技研究与发展》 * |
| 田久乐等: "基于同义词词林的词语相似度计算方法", 《吉林大学学报（信息科学版）》 * |
| 陈振亚等: "一种基于本体的文本特征选取方法", 《广西师范大学学报：自然科学版》 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102929931A (en) * | 2012-09-24 | 2013-02-13 | 上海师范大学 | Information credibility assessment method based on body in semantic net |
| CN109254959A (en) * | 2018-08-17 | 2019-01-22 | 广东技术师范学院 | A kind of data evaluation method, apparatus, terminal device and readable storage medium storing program for executing |
| CN109190342A (en) * | 2018-08-20 | 2019-01-11 | 济南大学 | The owner's auth method and community server of intelligence community |
| CN109190342B (en) * | 2018-08-20 | 2020-10-23 | 济南大学 | Owner identity verification method of smart community and community server |
| CN111815296A (en) * | 2020-07-23 | 2020-10-23 | 长沙公信诚丰信息技术服务有限公司 | Intelligent auditing method for third-party service system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN102681979B (en) | 2015-04-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Piao et al. | Privacy-preserving governmental data publishing: A fog-computing-based differential privacy approach | |
| Jin et al. | Application of data mining technology in financial risk analysis | |
| Lu et al. | Scalable news recommendation using multi-dimensional similarity and Jaccard–Kmeans clustering | |
| CN103207856A (en) | Ontology concept and hierarchical relation generation method | |
| CN103838857B (en) | Automatic service combination system and method based on semantics | |
| Zuo et al. | A large group decision-making method and its application to the evaluation of property perceived service quality | |
| CN109460460B (en) | Domain ontology construction method for intelligent application | |
| CN102681979A (en) | Content editing intelligent verifying method facing to open knowledge community | |
| CN110766272A (en) | Power business collaborative classification method and system based on ID3 decision tree algorithm | |
| Liu et al. | Cloud model-based PROMETHEE method under 2D uncertain linguistic environment | |
| CN109992725A (en) | A kind of social networks representation method based on two-way range internet startup disk | |
| CN110347401B (en) | API Framework service discovery method based on semantic similarity | |
| CN104679784A (en) | O2B intelligent searching method and system | |
| Han | System optimization of talent Life cycle management platform based on decision tree model | |
| Zhang et al. | Research on intelligent analysis of port logistics information based on dynamic data mining | |
| Chen et al. | Popular topic detection in Chinese micro-blog based on the modified LDA model | |
| Wu et al. | A novel topic clustering algorithm based on graph neural network for question topic diversity | |
| CN107153714B (en) | Service-relation-based change detection service chain on-demand generation method | |
| Li | Construction of an interactive sharing platform for competitive intelligence data of marine resources under the background of intelligence construction | |
| Wang et al. | ARIMA-BP integrated intelligent algorithm for China’s consumer price index forecasting and its applications | |
| Zhou | Application of-Means Clustering Algorithm in Energy Data Analysis | |
| Ma | The Research of Stock Predictive Model based on the Combination of CART and DBSCAN | |
| He et al. | Classification Algorithm based on Improved ID3 in Bank Loan Application | |
| Wang et al. | An ontology automation construction scheme for Chinese e‐government thesaurus optimizing | |
| Hao et al. | Ontology-based Software Trustworthy requirements and behavior modeling |
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 |