CN105975491A - Enterprise news analysis method and system - Google Patents
Enterprise news analysis method and system Download PDFInfo
- Publication number
- CN105975491A CN105975491A CN201610266431.1A CN201610266431A CN105975491A CN 105975491 A CN105975491 A CN 105975491A CN 201610266431 A CN201610266431 A CN 201610266431A CN 105975491 A CN105975491 A CN 105975491A
- Authority
- CN
- China
- Prior art keywords
- news
- association
- enterprise
- target enterprise
- full name
- 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.)
- Pending
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
- G06F16/951—Indexing; Web crawling techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
- G06Q40/03—Credit; Loans; Processing thereof
Abstract
The invention provides an enterprise news analysis method and system. The enterprise news analysis method comprises the following steps: judging and receiving the full name of a target enterprise; splitting the full name of the target enterprise to obtain the short name of the target enterprise; automatically searching associated news of the target enterprise in internet according to the full name and short name of the target enterprise; and analyzing the associated news by adopting an evaluation analysis method, and giving a credit evaluation to the target enterprise. According to the enterprise news analysis method and system, the users can rapidly and correctly obtain the credit evaluation of the target enterprise.
Description
Technical field
The present invention relates to big data analysis field, particularly relate to a kind of Company News and analyze method and system.
Background technology
In credit financing field, risk control personnel generally require and Target Enterprise carry out enterprise's public opinion image tune
Look into, thus this enterprise is provided objective evaluation.
Traditional forms of enterprises's investigation of public opinion method is by collecting enterprise's public opinion information artificially, and artificially checks public opinion
Information content, thus provide corresponding evaluation.But, the shortcoming of this traditional method is, to acquisition of information with
And the efficiency of information analysis is low, not only wastes time and energy, and occur that the probability that information is omitted is big.
Summary of the invention
The technical problem to be solved in the present invention is to provide a kind of Company News and analyzes method and system so that user
The credit rating to Target Enterprise can be drawn quickly and accurately.
For solving above-mentioned technical problem, present invention provide the technical scheme that
On the one hand, the present invention provides a kind of Company News to analyze method, is applied in computer, including: sentence
Break and receive Target Enterprise full name；Split Target Enterprise full name, obtain Target Enterprise and be called for short；Look forward to according to target
Industry full name and Target Enterprise are called for short, and automatically search the association news taking Target Enterprise in the Internet；Use and evaluate
Association news is analyzed by analysis method, and provides the credit rating to Target Enterprise.
Further, use web crawlers technology to search and take association news.
Further, before being analyzed step, also include: statistics Target Enterprise full name and Target Enterprise letter
Claim to associate the frequency of occurrences in news at each relatively low with the Target Enterprise degree of association, and it is low to will appear from frequency
It is judged to uncorrelated news in a range of association news and screens out.
Further, before being analyzed step, also include: calculate association news title between similar
Degree, and be judged as repeating news and screening out higher than a range of association news by similarity.
Further, evaluation analysis method includes positive negative sense evaluation analysis method and/or association personage and association enterprise
Industry analyzes method, and wherein, positive negative sense evaluation analysis method is: association news is used implicit expression semantic indexing method
Or keyword analysis method and keyword analysis carries out forward evaluation and negative sense evaluation to Target Enterprise；Association personage and affiliated enterprise
Analysis method is: extracts the organization's title in each association news and Personal name, and uses cluster
Method organization's title to being drawn into and Personal name cluster, by statistics Target Enterprise with each
Cluster simultaneously appear in frequency in relevant news, it is thus achieved that the association personage of Target Enterprise and affiliated enterprise,
Wherein, each classification that cluster is obtained represents an enterprise or individual, and organization's title includes enterprise
Industry full name or abbreviation.
Further, implicit expression semantic indexing method is, uses the positive negative sense grader pre-build to each pass
Connection news carries out positive negative sense evaluation.
Further, keyword analysis method and keyword analysis is, by extracting the forward key word in each association news
Word frequency and negative sense key word word frequency, and use the keywords database pre-build to compare, it is thus achieved that each pass
The positive negative sense evaluation of connection news.
On the other hand, the present invention provides a kind of Company News to analyze system, is applied in computer, including:
Enterprise name input module, it is judged that and receive Target Enterprise full name；Enterprise name splits module, will receive
Target Enterprise full name, split into Target Enterprise be called for short；Association news acquisition module, tears open according to enterprise name
Target Enterprise full name and Target Enterprise in sub-module are called for short, and automatically search the pass taking Target Enterprise in the Internet
Connection news；Association news analysis module, is carried out point the association news accessed by association news acquisition module
Analysis, and provide the credit rating to Target Enterprise.
Further, also including associating news screening module, association news is obtained by association news screening module
Association news obtained in module is screened.
Further, association news screening module includes that uncorrelated news screens out unit and duplicate removal unit, not phase
Closing news and screen out unit, statistics full name associates news with being called for short at each relatively low with the Target Enterprise degree of association
In the frequency of occurrences, and will appear from frequency and be judged to uncorrelated news and sieve less than a range of association news
Remove；Duplicate removal unit, calculates the similarity between the title of association news, and by similarity higher than certain limit
Association news be judged as repeating news screening out.
The Company News of inventive embodiments analyzes method and system, is applied in computer, by user is defeated
The Target Enterprise title needing inquiry entered is mated with the entry in enterprise name word library, is given and waits accordingly
Select enterprise's full name, such as, if user's input is the abbreviation of Target Enterprise, then can provide corresponding candidate enterprise
Industry full name, selects for user.After receiving Target Enterprise full name, according to certain rule, target will be looked forward to
Industry full name splits, thus obtains Target Enterprise and be called for short, such as, if Target Enterprise full name is " Hangzhou the
One Science and Technology Ltd. ", then " limited " and " company " common vocabulary of the two will the most screened fall, then
Remove place name " Hangzhou ", say, that referred to as " first science and technology " of the Target Enterprise finally obtained.With
Time use Target Enterprise full name and Target Enterprise to be called for short in internet environment, Target Enterprise situation to be searched
Rope, specifically, such as, for certain news item, if having mesh in this headline, summary and text
Mark enterprise's full name or abbreviation, then this news will be acquired.Use Target Enterprise full name and Target Enterprise simultaneously
Be called for short obtain be associated with Target Enterprise associate news, so that the association news searched for more is filled
Divide, comprehensively.Furthermore, it is necessary to explanation, by statistics Target Enterprise association news on the internet
Total amount and the source of association news, it can be estimated that the media exposure of this Target Enterprise and power of influence.Additionally,
For the relevant news obtained, use suitable evaluation analysis method to be analyzed, and provide one
Credit rating result to Target Enterprise more intuitively, so, can reduce one mesh of evaluation considerably
Time required for mark enterprise, it is also possible to be greatly saved manpower.
Therefore, the Company News that the present invention provides analyzes method and system, it is possible to make user quickly, accurately
Draw the credit rating to Target Enterprise.
Accompanying drawing explanation
Fig. 1 is the flow chart that the Company News that the embodiment of the present invention one provides analyzes method；
Fig. 2 is the block diagram that the Company News that the embodiment of the present invention two provides analyzes system；
Fig. 3 is the block diagram that the Company News that the embodiment of the present invention two provides analyzes system.
Detailed description of the invention
The present invention is further illustrated below by specific embodiment, it should be understood, however, that, these are implemented
Example is only used for specifically describing in more detail being used, and is not to be construed as limiting in any form this
Bright.
Embodiment one
In conjunction with Fig. 1, the embodiment of the present invention provides a kind of Company News to analyze method, is applied in computer,
The Company News of the embodiment of the present invention is analyzed the concrete steps of method and is included:
Step S101: judge and receive Target Enterprise full name；
Step S102: split Target Enterprise full name, obtains Target Enterprise and is called for short；
Step S103: be called for short according to Target Enterprise full name and Target Enterprise, automatically search in the Internet and take target
The association news of enterprise；
Step S104: use evaluation analysis method that association news is analyzed, and be given Target Enterprise
Credit rating.
The Company News of the embodiment of the present invention analyzes method, is applied in computer, by user inputted
The Target Enterprise title needing inquiry is mated with the entry in enterprise name word library, provides corresponding candidate and looks forward to
Industry full name, such as, if user's input is the abbreviation of Target Enterprise, then can provide corresponding candidate enterprise complete
Claim, select for user.After receiving Target Enterprise full name, by complete to Target Enterprise according to certain rule
Claim to split, thus obtain Target Enterprise and be called for short, such as, if Target Enterprise full name is " Hangzhou the first section
Skill company limited ", then " limited " and " company " common vocabulary of the two will the most screened fall, then remove
Place name " Hangzhou ", say, that referred to as " first science and technology " of the Target Enterprise finally obtained.Make simultaneously
It is called for short with Target Enterprise full name and Target Enterprise and in internet environment, Target Enterprise situation is scanned for, tool
Body ground, such as, for certain news item, if having Target Enterprise in this headline, summary and text
Full name or abbreviation, then this news will be acquired.Use Target Enterprise full name and Target Enterprise to be called for short to come simultaneously
What acquisition was associated with Target Enterprise associates news, so that the association news searched for is more fully, entirely
Face.Furthermore, it is necessary to explanation, by statistics Target Enterprise association news on the internet total amount and
The source of association news, it can be estimated that the media exposure of this Target Enterprise and power of influence.Additionally, for institute
The relevant news obtained, uses suitable evaluation analysis method to be analyzed, and it is straight to provide a comparison
The credit rating result to Target Enterprise seen, so, can reduce one Target Enterprise of evaluation considerably
The required time, it is also possible to be greatly saved manpower.
Therefore, the Company News that the embodiment of the present invention provides analyzes method and system, it is possible to make user quickly,
Draw the credit rating to Target Enterprise exactly.
Preferably, in step S103, use web crawlers technology to search and take association news.Due to web crawlers skill
The search effect of art is notable and application is ripe, uses web crawlers technology to obtain looking forward to target in the Internet
The news information that industry is associated, it can be ensured that acquired association news comprehensive.
It is further preferred that before being analyzed step, also include: statistics Target Enterprise full name and target enterprise
Industry abbreviation associates the frequency of occurrences in news at each relatively low with the Target Enterprise degree of association, and will appear from frequency
Rate is judged to uncorrelated news less than a range of association news and screens out.The acquisition of association news is to pass through
Target Enterprise full name and Target Enterprise abbreviation realize, and in the relevant news of institute got, exist and mesh
The association news (such as, the recruitment article of this Target Enterprise) that the mark enterprise degree of association is relatively low.For looking forward to target
The association news that the industry degree of association is relatively low, will add up Target Enterprise full name or Target Enterprise is called for short in this association news
The frequency of occurrences in summary, text, if the frequency of occurrences is less than a certain threshold value, then associates news identification by this
For uncorrelated news and screened out.Additionally, for recruitment class article, then be directed to headline and
" recruitment " in summary carries out keyword recognition, by this then news screen out.
It is further preferred that before being analyzed step, also include: calculate between the title of association news
Similarity, and be judged as repeating news and screening out higher than a range of association news by similarity.Due to mutually
Networking news often reprinted by media, cause obtained association news repeat more.The present embodiment
Be by judge the title similarity of relevant news the association news obtained is carried out duplicate removal.Specifically
Ground, first carries out participle to the title of every the association news obtained, headed for institute word segmentation result is converged
Always, formed in " word space ", and each title is converted into the vector in " word space ", additionally, for side
Just Similarity Measure, the term vector of each title will be normalized.More specifically, for any two
Corresponding to title two term vectors of bar association news, obtain by calculating the cosine between two term vectors
Obtain the similarity between them.For example it is assumed that association news A title corresponding to after normalization
Term vector is a, and the term vector after normalization that association is corresponding to news B is b, if vector a
And the cosine value between vector b is 1, now, association news A and the similarity assessment knot associating news B
Fruit is " identical "；If vector a and vector b between cosine value be 0, now, association news A with
The similarity assessment result of association news B is " entirely different ".It is to say, between vector a and vector b
Cosine value the least, then association news A the lowest with the similarity associated between news B.Preferably, this reality
Execute threshold value that example sets as (0,0.5], i.e. then think phase when the cosine value of two titles vectors exceedes this threshold value
Seemingly spend height, and regarded as repeating news, and it is bigger to select to retain power of influence in media corresponding to news
That association news.Furthermore, it is necessary to explanation, set threshold value not immobilizes, Ke Yigen
It is adjusted according to practical situation.
Furthermore, it is necessary to explanation, screen out the frequency of occurrences and be less than step and the sieve of a range of association news
Except similarity is higher than the step of a range of association news, the two step there is no sequencing, as long as
Carry out before being analyzed step.
It is further preferred that in step S104, evaluation analysis method include positive negative sense evaluation analysis method and/
Or association personage and affiliated enterprise analyze method, wherein, positive negative sense evaluation analysis method is: to association news
Use implicit expression semantic indexing method or keyword analysis method and keyword analysis that Target Enterprise carries out forward evaluation and negative sense evaluation；
Association personage and affiliated enterprise analyze method: use entity extraction algorithm to extract in each association news
Organization's title and Personal name, and use clustering method organization's title to being drawn into and a name
Claim to cluster, by statistics Target Enterprise and each cluster simultaneously appear in frequency in relevant news,
Obtaining association personage and the affiliated enterprise of Target Enterprise, wherein, each classification that cluster is obtained represents one
Individual enterprise or individual, and organization's title includes enterprise's full name or abbreviation.By every to Target Enterprise
Association news carries out positive negative sense evaluation analysis, it can be estimated that the media public praise of enterprise.By to Target Enterprise
Association news is associated personage and affiliated enterprise analyzes, and can obtain the relational network of Target Enterprise.
Specifically, implicit expression semantic indexing method is, uses the positive negative sense grader pre-build to each association
News carries out positive negative sense evaluation.Wherein it is desired to explanation, implicit expression semantic indexing is to be looked for by magnanimity document
Go out the association between vocabulary, when two vocabulary or one group of vocabulary occur in a large number in same document, the two word
Converge or this group vocabulary is judged as semantic relevant.The present embodiment uses a number of news sample in advance
Training is carried out by implicit expression semantic indexing model (Latent Semantic Indexing Model),
Thus obtain a positive negative sense grader.More specifically, in the classifier training stage, the present embodiment be by
All of news sample is used for LSI model training, thus obtains the conversion square from word frequency vector to Concept Vectors
Battle array.Afterwards, for every news item sample, will first obtain a word frequency vector, and by LSI model
Obtain a corresponding Concept Vectors, the seat that the Concept Vectors of every news is expressed as in concept space
Punctuate.By LSI model learning, it is possible to select n the concept dimension that can separate positive negative sample to greatest extent,
And choose the hyperplane being made up of in concept space this n concept dimension as forward negative sense grader.Right
The analysis phase of association news, the word frequency to all vocabulary in association news to be evaluated, form sparse word
Frequency vector, projects concept space by LSI model by sparse term vector, by forward negative sense grader i.e.
Can obtain associating the forward negative sense evaluation analysis result of news.
It should be noted that the abbreviation that LSI model is implicit expression semantic indexing model of the present embodiment, the two
Conceptive is equivalent.
It is further preferred that keyword analysis method and keyword analysis is, close by extracting the forward in each association news
Keyword word frequency and negative sense key word word frequency, and use the keywords database pre-build to compare, it is thus achieved that each
The positive negative sense evaluation of bar association news.Specifically, the keywords database in the present embodiment is, first part forward
Extract key word in news sample and part negative sense news sample, and these key words passing through to be extracted build
About the basic keywords database of positively and negatively vocabulary, additionally, also this basic keywords database is expanded,
Concrete extending method is, with Google increase income bag word2vec method for algorithm basis, utilize gather also
Treated news sample is that keywords database is expanded by material, thus obtains the key word of the present embodiment
Storehouse.In the analysis phase to association news, when association news to be evaluated is processed, just extracting respectively
To key word word frequency and negative sense key word word frequency, obtain by key word word frequency is compared with keywords database
The positive negative sense evaluation analysis result of each association news.
Embodiment two
Shown in Fig. 2, the present embodiment provides a kind of Company News to analyze system, is applied in computer,
The Company News of the present embodiment is analyzed system and is included: enterprise name input module 1, it is judged that and receive target enterprise
Industry full name；Enterprise name splits module 2, by complete for the Target Enterprise received in enterprise name input module 1
Claim, split into Target Enterprise and be called for short；Association news acquisition module 3, splits in module 2 according to enterprise name
Target Enterprise full name and Target Enterprise be called for short, in the Internet, automatically search the association news taking Target Enterprise；
Association news analysis module 4, is analyzed the association news accessed by association news acquisition module 3,
And provide the credit rating to Target Enterprise.
The Company News of the present invention analyzes system, is applied in computer, by being looked into by the needs that user inputs
The Target Enterprise title ask is mated with the entry in enterprise name word library, provides corresponding candidate enterprise full name,
Such as, if user's input is the abbreviation of Target Enterprise, then corresponding candidate enterprise full name can be provided, for
Family selects.After receiving Target Enterprise full name, according to certain rule, Target Enterprise full name will be torn open
Point, thus obtain Target Enterprise and be called for short, such as, if Target Enterprise full name is " the scientific and technological limited public affairs in Hangzhou first
Department ", then " limited " and " company " the two common vocabulary general the most screened fall, then remove place name " Hangzhoupro
State ", say, that referred to as " first science and technology " of the Target Enterprise finally obtained.Use target enterprise simultaneously
Target Enterprise situation is scanned in internet environment by industry full name and Target Enterprise abbreviation, specifically, and example
As, for certain news item, if this headline, summary and text have Target Enterprise full name or letter
Claim, then this news will be acquired.Use Target Enterprise full name and Target Enterprise to be called for short to obtain and mesh simultaneously
The association news that mark enterprise is associated, so that the association news searched for is more fully, comprehensively.Additionally,
For the relevant news obtained, use suitable evaluation analysis method to be analyzed, and provide one
Credit rating result to Target Enterprise more intuitively, so, can reduce one mesh of evaluation considerably
Time required for mark enterprise, it is also possible to be greatly saved manpower.
Therefore, the Company News that the present invention provides analyzes method and system, it is possible to make user quickly, accurately
Draw the credit rating to Target Enterprise.
Preferably, also include associating news screening module 5, as it is shown on figure 3, association news screening module 5
Association news obtained in association news acquisition module 3 is screened.
It is further preferred that association news screening module 5 includes that uncorrelated news screens out unit and duplicate removal unit,
Uncorrelated news screens out unit, and statistics full name associates at each relatively low with the Target Enterprise degree of association with being called for short
The frequency of occurrences in news, and will appear from frequency and be judged to uncorrelated news less than a range of association news
And screen out；Duplicate removal unit, calculates the similarity between the title of association news, and by similarity higher than certain
The association news of scope is judged as repeating news and screening out.
Although present invention has been a certain degree of description, it will be apparent that, without departing from the present invention spirit and
Under conditions of scope, the suitable change of each condition can be carried out.It is appreciated that and the invention is not restricted to described reality
Executing scheme, and be attributed to the scope of claim, it includes the equivalent of described each factor.
Claims (10)
1. Company News analyzes a method, is applied in computer, it is characterised in that described method includes:
Judge and receive Target Enterprise full name；
Split described Target Enterprise full name, obtain Target Enterprise and be called for short；
It is called for short according to described Target Enterprise full name and described Target Enterprise, automatically searches in the Internet and take target enterprise
The association news of industry；
Use evaluation analysis method that described association news is analyzed, and provide the letter to described Target Enterprise
Reputation is evaluated.
Company News the most according to claim 1 analyzes method, it is characterised in that described in search and take target
In the association news step of enterprise, use web crawlers technology to search and take described association news.
Company News the most according to claim 1 analyze method, it is characterised in that described in be analyzed
Before step, also include: add up described Target Enterprise full name and described Target Enterprise be called for short with described target
The frequency of occurrences in described each the association news that enterprise's degree of association is relatively low, and will appear from frequency less than certain
The described association news of scope is judged to uncorrelated news and screens out.
Company News the most according to claim 1 analyze method, it is characterised in that described in be analyzed
Before step, also include: calculate the similarity between the title of described association news, and similarity is higher than
A range of described association news is judged as repeating news and screening out.
Company News the most according to any one of claim 1 to 4 analyzes method, it is characterised in that
Described being analyzed in step, described evaluation analysis method includes positive negative sense evaluation analysis method and/or pass
Connection personage and affiliated enterprise's analysis method, wherein,
Described positive negative sense evaluation analysis method is: described association news is used implicit expression semantic indexing method or key
Word is analyzed method and described Target Enterprise is carried out forward evaluation and negative sense evaluation；
Described association personage and affiliated enterprise analyze method: extract the tissue in described each association news
Organization names and Personal name, and use the clustering method described organization title to being drawn into and a name
Claim to cluster, simultaneously appear in described relevant news by adding up described Target Enterprise with each cluster
In frequency, it is thus achieved that the association personage of described Target Enterprise and affiliated enterprise, wherein, cluster obtained every
One classification represents an enterprise or individual, and described organization title includes enterprise's full name or abbreviation.
Company News the most according to claim 5 analyzes method, it is characterised in that described implicit expression is semantic
Indexing method is, uses the positive negative sense grader pre-build that described each association news is carried out positive negative sense and comments
Valency.
Company News the most according to claim 5 analyzes method, it is characterised in that described key word divides
Analysis method is, by extracting the forward key word word frequency in described each association news and negative sense key word word
Frequently, and use the keywords database that pre-builds to compare, it is thus achieved that the positive negative sense of described each association news
Evaluate.
8. Company News analyzes a system, is applied in computer, it is characterised in that this system includes:
Enterprise name input module: judge and receive Target Enterprise full name；
Enterprise name fractionation module: the Target Enterprise full name that will receive in described enterprise name input module,
Split into Target Enterprise to be called for short；
Association news acquisition module: according to described enterprise name split the described Target Enterprise full name in module and
Described Target Enterprise is called for short, and automatically searches the association news taking Target Enterprise in the Internet；
Association news analysis module: the described association news accessed by described association news acquisition module is entered
Row is analyzed, and provides the credit rating to Target Enterprise.
Company News the most according to claim 8 analyzes system, it is characterised in that also include that association is new
Hearing screening module, described association news screening module is to the association obtained in described association news acquisition module
News is screened.
Company News the most according to claim 9 analyzes system, it is characterised in that described association is new
Hear screening module and include that uncorrelated news screens out unit and duplicate removal unit；
Described uncorrelated news screens out unit, adds up described Target Enterprise full name and described Target Enterprise is called for short
Described each the frequency of occurrences that associate in news relatively low with the described Target Enterprise degree of association, and by described go out
Existing frequency is judged to uncorrelated news less than a range of described association news and screens out；
Described duplicate removal unit, calculates the similarity between the title of described association news, and by described similarity
It is judged as repeating news and screening out higher than a range of described association news.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610266431.1A CN105975491A (en) | 2016-04-26 | 2016-04-26 | Enterprise news analysis method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610266431.1A CN105975491A (en) | 2016-04-26 | 2016-04-26 | Enterprise news analysis method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN105975491A true CN105975491A (en) | 2016-09-28 |
Family
ID=56993118
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610266431.1A Pending CN105975491A (en) | 2016-04-26 | 2016-04-26 | Enterprise news analysis method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN105975491A (en) |
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107291812A (en) * | 2017-05-18 | 2017-10-24 | 辛柯俊 | It is a kind of that approximating method and system are learnt to expert naturally semantic in company information |
| CN107688564A (en) * | 2017-08-31 | 2018-02-13 | 平安科技（深圳）有限公司 | Subject of news Corporate Identity method, electronic equipment and computer-readable recording medium |
| CN107909274A (en) * | 2017-11-17 | 2018-04-13 | 平安科技（深圳）有限公司 | Enterprise investment methods of risk assessment, device and storage medium |
| CN108021596A (en) * | 2016-11-04 | 2018-05-11 | 百度在线网络技术（北京）有限公司 | The methods of exhibiting and device of news |
| CN108460014A (en) * | 2018-02-07 | 2018-08-28 | 百度在线网络技术（北京）有限公司 | Recognition methods, device, computer equipment and the storage medium of business entity |
| CN108460016A (en) * | 2018-02-09 | 2018-08-28 | 中云开源数据技术（上海）有限公司 | A kind of entity name analysis recognition method |
| CN108846547A (en) * | 2018-05-06 | 2018-11-20 | 成都信息工程大学 | A kind of Enterprise Credit Risk Evaluation method of dynamic adjustment |
| CN109598393A (en) * | 2017-09-30 | 2019-04-09 | 北京国双科技有限公司 | A kind of analysis method and device of the influence information that event generates enterprise |
| CN109670837A (en) * | 2018-11-30 | 2019-04-23 | 平安科技（深圳）有限公司 | Recognition methods, device, computer equipment and the storage medium of bond default risk |
| CN109902908A (en) * | 2018-10-18 | 2019-06-18 | 国家电网公司 | A kind of enterprise's method for quantitatively evaluating based on public data |
| CN110020048A (en) * | 2017-10-27 | 2019-07-16 | 北京宸信征信有限公司 | A kind of business risk evaluation system and method based on open source data |
| CN110442704A (en) * | 2019-08-13 | 2019-11-12 | 重庆誉存大数据科技有限公司 | A kind of Company News screening technique and system |
| WO2020147599A1 (en) * | 2019-01-14 | 2020-07-23 | 紫光云数科技有限公司 | Comprehensive information acquisition system and method and slave acquisition node system |
| CN112613299A (en) * | 2020-12-25 | 2021-04-06 | 北京知因智慧科技有限公司 | Method and device for constructing enterprise synonym library and electronic equipment |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101751458A (en) * | 2009-12-31 | 2010-06-23 | 暨南大学 | Network public sentiment monitoring system and method |
-
2016
- 2016-04-26 CN CN201610266431.1A patent/CN105975491A/en active Pending
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101751458A (en) * | 2009-12-31 | 2010-06-23 | 暨南大学 | Network public sentiment monitoring system and method |
Non-Patent Citations (4)
| Title |
|---|
| 候小可: "微博新闻话题的情感分析研究", 《中国优秀硕士学位论文全文数据库信息科技辑》 * |
| 孙丽萍等: "基于构成模式和条件随机场的企业简称预测", 《计算机应用》 * |
| 谭征等: "中文评教文本分类模型的研究", 《烟台大学学报(自然科学与工程版)》 * |
| 贾超: "基于情感分析与垂直搜索引擎的新闻搜索监控系统设计与研究", 《万方数据库》 * |
Cited By (18)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108021596A (en) * | 2016-11-04 | 2018-05-11 | 百度在线网络技术（北京）有限公司 | The methods of exhibiting and device of news |
| CN108021596B (en) * | 2016-11-04 | 2021-12-28 | 百度在线网络技术（北京）有限公司 | News display method and device |
| CN107291812A (en) * | 2017-05-18 | 2017-10-24 | 辛柯俊 | It is a kind of that approximating method and system are learnt to expert naturally semantic in company information |
| CN107688564A (en) * | 2017-08-31 | 2018-02-13 | 平安科技（深圳）有限公司 | Subject of news Corporate Identity method, electronic equipment and computer-readable recording medium |
| CN109598393A (en) * | 2017-09-30 | 2019-04-09 | 北京国双科技有限公司 | A kind of analysis method and device of the influence information that event generates enterprise |
| CN110020048B (en) * | 2017-10-27 | 2021-09-14 | 北京宸信征信有限公司 | Enterprise risk evaluation system and method based on open source data |
| CN110020048A (en) * | 2017-10-27 | 2019-07-16 | 北京宸信征信有限公司 | A kind of business risk evaluation system and method based on open source data |
| CN107909274A (en) * | 2017-11-17 | 2018-04-13 | 平安科技（深圳）有限公司 | Enterprise investment methods of risk assessment, device and storage medium |
| CN107909274B (en) * | 2017-11-17 | 2023-02-28 | 平安科技（深圳）有限公司 | Enterprise investment risk assessment method and device and storage medium |
| CN108460014B (en) * | 2018-02-07 | 2022-02-25 | 百度在线网络技术（北京）有限公司 | Enterprise entity identification method and device, computer equipment and storage medium |
| CN108460014A (en) * | 2018-02-07 | 2018-08-28 | 百度在线网络技术（北京）有限公司 | Recognition methods, device, computer equipment and the storage medium of business entity |
| CN108460016A (en) * | 2018-02-09 | 2018-08-28 | 中云开源数据技术（上海）有限公司 | A kind of entity name analysis recognition method |
| CN108846547A (en) * | 2018-05-06 | 2018-11-20 | 成都信息工程大学 | A kind of Enterprise Credit Risk Evaluation method of dynamic adjustment |
| CN109902908A (en) * | 2018-10-18 | 2019-06-18 | 国家电网公司 | A kind of enterprise's method for quantitatively evaluating based on public data |
| CN109670837A (en) * | 2018-11-30 | 2019-04-23 | 平安科技（深圳）有限公司 | Recognition methods, device, computer equipment and the storage medium of bond default risk |
| WO2020147599A1 (en) * | 2019-01-14 | 2020-07-23 | 紫光云数科技有限公司 | Comprehensive information acquisition system and method and slave acquisition node system |
| CN110442704A (en) * | 2019-08-13 | 2019-11-12 | 重庆誉存大数据科技有限公司 | A kind of Company News screening technique and system |
| CN112613299A (en) * | 2020-12-25 | 2021-04-06 | 北京知因智慧科技有限公司 | Method and device for constructing enterprise synonym library and electronic equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN105975491A (en) | Enterprise news analysis method and system | |
| CN109791569B (en) | Causal relationship identification device and storage medium | |
| CN107577739B (en) | Semi-supervised domain word mining and classifying method and equipment | |
| US10387805B2 (en) | System and method for ranking news feeds | |
| CN107301171A (en) | A kind of text emotion analysis method and system learnt based on sentiment dictionary | |
| CN106021410A (en) | Source code annotation quality evaluation method based on machine learning | |
| US20170091318A1 (en) | Apparatus and method for extracting keywords from a single document | |
| Liliana et al. | Indonesian news classification using support vector machine | |
| CN113255340B (en) | Theme extraction method and device for scientific and technological requirements and storage medium | |
| Vainshtein et al. | A hybrid approach for automatic model recommendation | |
| AU2013365452A1 (en) | Document classification device and program | |
| CN105550168A (en) | Method and device for determining notional words of objects | |
| CN110910175A (en) | Tourist ticket product portrait generation method | |
| Nguyen et al. | An ensemble of shallow and deep learning algorithms for Vietnamese sentiment analysis | |
| Jaman et al. | Sentiment analysis of customers on utilizing online motorcycle taxi service at twitter with the support vector machine | |
| CN115238154A (en) | Search engine optimization system | |
| CN109582743B (en) | Data mining system for terrorist attack event | |
| Zhang et al. | A hot spot clustering method based on improved kmeans algorithm | |
| Saha et al. | A large scale study of SVM based methods for abstract screening in systematic reviews | |
| Aissa et al. | Using machine learning to predict outcomes of accident cases in moroccan courts | |
| CN107544962A (en) | Social media text query extended method based on Similar Text feedback | |
| Hartanto et al. | Classifying User Personality Based on Media Social Posts Using Support Vector Machine Algorithm Based on DISC Approach | |
| CN111538898A (en) | Web service package recommendation method and system based on combined feature extraction | |
| Swezey et al. | Automatic detection of news articles of interest to regional communities | |
| US20090254581A1 (en) | Knowledge discovery system capable of custom configuration by multiple users |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20160928 |