CN110020048A

CN110020048A - A kind of business risk evaluation system and method based on open source data

Info

Publication number: CN110020048A
Application number: CN201711022805.6A
Authority: CN
Inventors: 张守义
Original assignee: Beijing Chen Xin Credit Investigation Co Ltd
Current assignee: Chongqing Chenyu Information Technology Co.,Ltd.
Priority date: 2017-10-27
Filing date: 2017-10-27
Publication date: 2019-07-16
Anticipated expiration: 2037-10-27
Also published as: CN110020048B

Abstract

The invention discloses a kind of business risk evaluation systems and method based on open source data；The system includes that data crawl module, and data relevant to enterprise to be evaluated can be crawled from webpage, and data word segmentation module can do word segmentation processing to the data crawled, and count word frequency；It further include grading module, its participle and its word frequency for being used to be obtained according to word segmentation processing are judged, wherein grading module includes multiple submodule, each submodule makes evaluation for the factor for the one aspect for influencing enterprise development prospect, finally obtain reasonable enterprise's scoring, the risk that enterprise can be judged according to score height, can also judge to compare according to the difference of the scoring of different enterprises.

Description

A kind of business risk evaluation system and method based on open source data

Technical field

The present invention relates to a kind of enterprise data analysis processing system more particularly to a kind of business risks based on open source data Evaluation system and method.

Background technique

With the arriving of big data era, people increasingly pay attention to judging by data analysis or handling more intractable Problem, wherein simple statistics, adduction be still easier to understand and handle, if it is known that data and the result of needs it Between corresponding relationship when being not obvious, generally require to carry out statistical disposition by special program or device, but specific such as where Reason, is calculated by which type of computing device, rare in the prior art to be related to；

Specifically, in modern society enterprise, company substantial amounts, wherein each enterprise-like corporation's quality is very different, Before doing the decisions such as selection affiliate, the selection investment objective, it is necessary to fully understand the potentiality of the enterprise, ability etc. The case where aspect, it is often more important that need lateral comparison, need to filter out in numerous type of industry it is most suitable, can Meet the enterprise of self-demand, various information are countless on present network, are in general difficult to screen in the network information of magnanimity Valuable information out, and being comprehensively compared, and even if statistical information and being analyzed, spent by time cost manpower Cost be also it is huge, be often possible to lose more than gain, so needing one kind that can fast, accurately and comprehensively assess, analyze enterprise The system or method of risk, to meet the demand.

Summary of the invention

In order to overcome the above problem, present inventor has performed sharp studies, design a kind of enterprise based on open source data Risk Evaluating System and method；The system can crawl out article relevant to specified enterprise etc. in many and diverse network of big information Information, and it is divided into multiple big classifications with business risk related content for various involved in the information, and under big classification Multiple small classifications are set, are assessed respectively, and then are obtained more scientific reasonable as a result, due to considering after comprehensive assessment Factor it is comprehensive enough, finally obtained result is naturally more reasonable, accurate, and this evaluation process is more rapid, can Meets the needs of all kinds of users, wherein the system includes that data crawl module, can be crawled from webpage and enterprise to be evaluated The relevant data of industry, data word segmentation module can do word segmentation processing to the data crawled, and count word frequency；It further include scoring mould Block, the participle and its word frequency for being used to be obtained according to word segmentation processing are judged that wherein grading module includes multiple submodule, often One submodule makes evaluation for the factor for the one aspect for influencing enterprise development prospect, finally obtains reasonable enterprise and comments Point, the risk of enterprise can be judged according to score height, can also judge to compare according to the difference of the scoring of different enterprises, from And complete the present invention.

In particular it is object of the present invention to provide a kind of business risk evaluation system based on open source data, this is System includes

Data crawl module 1, are used to crawl data from webpage,

Data word segmentation module 2 is used to crawl data the data text that module 1 crawls and does word segmentation processing, and counts Word frequency；With

Grading module 3 is used to provide enterprise's scoring according to the word frequency of participle.

Beneficial effect possessed by the present invention includes:

It is each that the business risk evaluation system based on open source data provided according to the present invention can obtain in all directions enterprise The information of a aspect, and to reasonable scoring is provided by the grading module and keyword of setting, it can obtain in a short time It is evaluated to business risk, knows business risk, and system tool is commented there are four the submodule that scores from many aspects respectively Point, the factor that when scoring considers is sufficient, and appraisal result is more scientific rationally.

Detailed description of the invention

The business risk evaluation system based on open source data that Fig. 1 shows a kind of preferred embodiment according to the present invention is whole Structural schematic diagram.

Drawing reference numeral explanation:

1- data crawl module

2- data word segmentation module

3- grading module

11- input unit

31- enterprise operation and management scoring submodule

32- enterprise competitiveness scoring submodule

33- enterprise development prospect scoring submodule

34- industry development environment scoring submodule

311- enterprise key person index dimension judging part

312- corporate social reputation index dimension judging part

313- public records index dimension judging part

The horizontal index dimension judging part of 321- enterprise innovation

322- brand influence index dimension judging part

331- enterprise investment and financing index dimension judging part

332- product renewing iteration index dimension judging part

333- product life cycle index dimension judging part

334- capital market dynamic indicator dimension judging part

341- industry prospect index dimension judging part

342- national policy index dimension judging part

Specific embodiment

Below by drawings and examples, the present invention is described in more detail.Illustrated by these, the features of the present invention It will be become more apparent from advantage clear.

Dedicated word " exemplary " means " being used as example, embodiment or illustrative " herein.Here as " exemplary " Illustrated any embodiment should not necessarily be construed as preferred or advantageous over other embodiments.Although each of embodiment is shown in the attached drawings In terms of kind, but unless otherwise indicated, it is not necessary to attached drawing drawn to scale.

A kind of business risk evaluation system based on open source data provided according to the present invention, as shown in fig. 1, the system Module 1 is crawled including data, is used to crawl data from webpage, it is preferable that the data crawl module 1 from the webpage of open source Data are obtained, including information disclosed in all kinds of articles delivered, news report and each government bodies etc.；

It is further preferred that the data crawl in module 1 can also external input device 11, such as keyboard, mouse lead to Cross the input unit 11 input retrieval information, such as enterprise name, it is described crawl module 1 crawl containing/refer to the enterprise name Article.

It includes crawling engine, crawler module, downloading middleware, download module, crawler middleware that the data, which crawl module 1, With element pipeline；

Wherein, specifically, the data, which crawl module 1 and crawl the process of web data, includes the following steps:

Step 1, it crawls engine and obtains initial request from crawler module,

Step 2, it crawls engine and the request obtained from crawler module is included in task scheduling,

Step 3, the next request of task scheduling return is crawled engine,

Step 4, it crawls engine and the request that task scheduling returns is sent to download module by downloading middleware,

Step 5, download module downloads the page, and a response can be generated when download module has downloaded the page, and under passing through Load middleware, which is sent to, crawls engine,

Step 6, after crawling the response that engine receives download module transmission, crawler module is sent to by crawler middleware,

Step 7, it after crawler resume module crawls the response that engine is sent, is climbed by crawler middleware to engine return is crawled Element and new request are taken,

Step 8, it crawls engine and the processed element that crawls is sent to element pipeline, then send processing request to task Plan and wait next possible request,

Step 9, repeat the above steps 1-8, until task scheduling not new request.

The system further includes data word segmentation module 2, is used to crawl data the data text that module 1 crawls and segments Processing, the word segmentation processing, which refers to, splits into multiple phrase/participles for article, or extract from article multiple phrases/point Word, and all phrase/participles are condensed together, obtain the frequency of occurrence of each participle, the as word frequency of the participle.

The treatment process of heretofore described word segmentation processing includes the following steps:

Step 1, increase DOCID unique identification to the open article crawled from network,

Step 2, the article crawled is handled, article information is divided into article essential information and article segments two major classes, respectively Summarizing,

Step 3, by treated, data are stored in database,

Step 4, the word frequency of each participle is counted.

Wherein, article essential information includes: DOCID, title, chained address, author, issuing time, acquisition time and text Chapter keyword.

In the present invention, can there is no special provision to this using conventional participle mode.It is preferred that the participle Processing can be handled by the semantic open platform of Chinese of Shanghai Bo Sen data technologies Co., Ltd, obtain the article Essential information.

In one preferred embodiment, sentiment analysis also is done to the article crawled during the word segmentation processing, The sentiment analysis refers to that analysis obtains the non-negative probability and negative probability of this article, and the non-negative probability and negative The two values of probability be added and be 1.For example, by the semantic open platform of Chinese of Shanghai Bo Sen data technologies Co., Ltd, It can be obtained non-negative probability and negative probability in word segmentation processing.

By non-negative probability and negative probability, know that this article is positive propaganda, negative campaigning or neutral publicity.Tool Body, judge this article for neutrality publicity when the difference of non-negative probability and negative probability is between -0.1 and 0.1；When non-negative Judge that this article is positive propaganda when the numerical value that the difference of face probability and negative probability is 0.1 or more；When non-negative probability and negatively The difference of probability judges that this article is negative campaigning when being -0.1 numerical value below.

For plurality of articles, obtaining every article with such as above method respectively is positive propaganda, negative campaigning or neutrality Publicity.Then, the article quantity and negative campaigning article quantity for counting positive propaganda, calculate the article quantity of positive propaganda The sum of ratio with the sum of negative campaigning article quantity, when the ratio is 2 or more, the final scoring of the said firm is additional to be increased 5 or 10 points, when the ratio is below 0.5, the final scoring of the said firm plus it is additional increase by -5 or -10 points, when the ratio between When 0.5~2, extra process is not done for finally scoring；Wherein it is preferred to when the ratio 2 more than and less than 5 when, the public affairs The final scoring of department is additional to increase by 5 points, and when the ratio is 5 or more, the final scoring of the said firm is additional to increase by 10 points；Work as institute When stating ratio below 0.5 and being greater than 0.2, the final scoring of the said firm is additional to increase -5 points, when the ratio is below 0.2 When, the final scoring of the said firm is additional to increase -10 points.

The system further includes grading module 3, is used to provide enterprise's scoring according to the word frequency of participle；

Preferably, institute's scoring module 3 is analyzed respectively for many aspects of enterprise, and according to predefined weight coefficient It sums up, and then obtains final overall score.

Further, before institute's scoring module 3 is for the enterprise operation and management of enterprise, enterprise competitiveness, enterprise development Four aspects of scape and industry development environment are analyzed；Specifically, institute's scoring module 3 includes enterprise operation and management scoring Module 31, enterprise competitiveness scoring submodule 32, enterprise development prospect scoring submodule 33 and industry development environment scoring Module 34 calculates the scoring of various aspects according to each submodule respectively, obtains the scoring of each submodule, weighs according still further to major class It sums up to obtain the final scoring of the enterprise again；Enterprise's wind can be judged by the scoring event of each enterprise of lateral comparison The size of danger；Preferably, score is higher, and the situation of enterprises is better, and risk is smaller；

Preferably, the major class weight coefficient of the enterprise operation and management scoring submodule 31 is 0.4, and enterprise competitiveness is commented The major class weight coefficient of molecular modules 32 is 0.2, and the major class weight coefficient of enterprise development prospect scoring submodule 33 is 0.1, row The major class weight coefficient of industry development environment scoring submodule 34 is 0.3；The corresponding major class weight of the scoring of each submodule Summation after multiplication is exactly the overall score.

During actual risk assessment, general enterprise be all it is more advantageous in some aspects, in some aspects Some are insufficient, but how to balance these advantages and deficiency, are always a problem, in order to solve this problem, people are often Research and analysis can be distinguished to each single item, can spend very big energy and time in this process, and different mechanism, people The research of member, analysis characteristic are also not quite similar, and eventually lead to and are difficult to properly be fused together, not only waste the time, but also be difficult to Good desired effect is obtained, so aforementioned four scoring submodule is just arranged in the present invention, can either ensure to score is reasonable And accuracy, it can also seek unity of standard, improve efficiency, can rapidly and accurately know relevant information, and be convenient for lateral comparison, Convenient for laterally obtaining risk assessment by comparing the score value between similar enterprises.

Several index dimensions will be investigated in each submodule, i.e., in each submodule include two with Upper index dimension judging part all includes more than two index item in each index dimension judging part, and distinguishes each index item It scores, the subclass weight phase after the scoring of each index item in an index dimension judging part is added with the index dimension Multiply, obtains the scoring of the index dimension, each index dimension scoring and the as submodule scoring.Wherein, each index Dimension is all corresponding with a subclass weight.

Specifically, each index dimension includes more than two index item, is all stored in each index item There is more than one index keyword, and be based on the index keyword, screening one by one is carried out in the participle extracted, knows institute It states in participle and which index keyword is contained, that is, find out participle identical with index keyword, and know that the participle/index is closed The word frequency of key word, and then obtain the scoring of the index item.

Heretofore described index keyword be all analyzed comprehensively by applicant, repeatedly attempt obtained from it is optimal most Suitable vocabulary, the index keyword are all representativenesses, are easy to split, the word of few ambiguity, but also be frequent in webpage report Use, and word frequency higher word big with index relevance.

Preferably, the word frequency sum of all index keyword correspondence/hit participles is total word frequency in index item, is being divided The corresponding word frequency of undiscovered index keyword is zero in word, is also stored with judgment module in each index item, institute State the scoring that judgment module judges each index item according to the content of total word frequency or hit keyword.

It is further preferred that the scoring in the Rule of judgment is the score between -100~100, for specific targets key Word, corresponding scoring can be positive or negative.

In one preferred embodiment, enterprise operation and management scoring submodule 31 includes enterprise's key person index dimension Judging part 311, corporate social reputation index dimension judging part 312 and public records index dimension judging part 313；The enterprise closes The subclass weight of key people's index dimension is 0.3, and the subclass weight of the corporate social reputation index dimension is 0.3, described public The subclass weight of record index dimension is 0.4.

Wherein, enterprise's key person index dimension judging part 311 includes:

Dong supervises high internet exposure index item, index keyword therein include summit, forum, annual meeting, innovation conference, Special visit, seminar, develops conference at product news conference；

Social responsibility index item, index keyword therein include public good, charitable, industry leader, model worker；

Regional telephone distribution index item, index keyword therein include leave office, resign, is negative, disturbance, fail, is undisciplined, It looked into, investigated；With

Positive news information index item, index keyword therein include Outstanding Contribution Award, man of the hour, outstanding person, Leader, best senior executive, annual character.

The corporate social reputation index dimension judging part 312 includes:

Winning information index item, index keyword therein include prize-winning enterprise, silver medal, gold medal, Outstanding Contribution Award, especially Encourage Innovation Awards, outstanding operation prize, medal, honorary certificate, prize-giving grand ceremony；

Information index item is commended, index keyword therein includes medal, honorary certificate, favorable comment；

Reputation reputation index item, index keyword therein include well received, favorable comment, phenomenon grade, public praise prize, it is best, Enterprise's public praise, internet word-of-mouth；

Commonweal information index item, index keyword therein include public good Contribution Prize, charitable, donations, donation, public good work Dynamic, utility；

Business ethics good index item, index keyword therein include business ethics enterprise, moral enterprise, and

Business ethics ruins index item, and index keyword therein includes that business ethics is ruined, ruined.

The public records index dimension judging part 313 includes:

Administrative penalty information index item, index keyword therein include administrative penalty, responsibility dispute, illegal operation, relate to Dislike violation, punishment publicity；

Administrative permission information index item, index keyword therein include operation permission, administrative permission, licensing；

Exception information index item is managed, index keyword therein manages register, exception including the abnormal register of operation, exception Distributors manage abnormal enterprise；

Tax negative information index item, index keyword therein are different including exception of paying taxes, tax evasion, tax declaration Often；

Media negative information index item, index keyword therein include being accused of, substandard product, recalling, rectify and improve, working Dispute negative press, reduces the staff, runs away, faking；With

Persecutio information index item, index keyword therein include encroach right, lose a lawsuit, prosecuting, lawsuit.

Preferably, Dong supervises in high internet exposure index item and social responsibility index item judgment module judgment rule Be: total word frequency is that 0 news commentary is divided into 0 point, when total word frequency is between (0-3), and scoring is 30 points, total word frequency between [3-5) when, scoring is 50 points, total word frequency between [5- ∞) when, scoring is 100 points；

Judgment module judgment rule is all in winning information index item, commendation information index item and commonweal information index item: Total word frequency is that 0 news commentary is divided into 0 point, when total word frequency is between (0-3), and scoring is 10 points, total word frequency between [3-5) when, scoring is 25 Point, total word frequency between [5-7) when, scoring is 50 points, total word frequency between [7-10) when, scoring is 75 points, and total word frequency is between [10- When ∞), scoring is 100 points；

Positive news information index item, business ethics good index item, reputation reputation index item and administrative permission information refer to Judgment module judgment rule is all in mark item: total word frequency is that 0 news commentary is divided into 0 point, when total word frequency is between (0-5), and scoring is 10 points, Total word frequency between [5-10) when, scoring is 25 points, total word frequency between [10-15) when, scoring is 50 points, and total word frequency is between [15- 20) when, scoring is 75 points, total word frequency between [20- ∞) when, scoring is 100 points；

Regional telephone distribution index item, business ethics ruin index item, administrative penalty information index item, manage exception information Judgment module is sentenced in index item, tax negative information index item, media negative information index item and persecutio information index item Disconnected rule is all: total word frequency is that 0 news commentary is divided into 0 point, when total word frequency is between (0-5), and scoring is -10 points, and total word frequency is between [5- 10) when, scoring is -25 points, total word frequency between [10-15) when, scoring is -50 points, total word frequency between [15-20) when, score for - 75 points, total word frequency between [20- ∞) when, scoring is -100 points.

In the application bracket " [" indicate to include the numerical value, round bracket " (" and ") " expression does not include the numerical value, such as [5-10) it indicates to be more than or equal to 5 and less than 10.

In one preferred embodiment, enterprise competitiveness scoring submodule 32 includes the horizontal index dimension of enterprise innovation Spend judging part 321 and brand influence index dimension judging part 322；The subclass weight of the horizontal index dimension of enterprise innovation is 0.5, the subclass weight of the brand influence index dimension is 0.5.

Wherein, the horizontal index dimension judging part 321 of the enterprise innovation includes:

Patent application index item, index keyword therein include patent, patent of invention, patent certificate；

Trade mark registration index item, index keyword therein include trade mark, trademark application；

Copyright delivers index item, and index keyword therein includes copyright, copyright.

The brand influence index dimension judging part 322 includes:

Brand recognition index item, index keyword therein include popularity, esbablished corporation, well-known trademark, inspection-free production Product, reputation；

Brand share index item, index keyword therein include occupation rate of market, monopolization.

It is accounted for wherein it is preferred to which trade mark registration index item, copyright deliver index item, brand recognition index item and brand Judgment module judgment rule is all in rate index item: total word frequency is that 0 news commentary is divided into 0 point, when total word frequency is between (0-3), scoring Be 10 points, total word frequency between [3-5) when, scoring is 25 points, total word frequency between [5-7) when, scoring is 50 points, total word frequency between [7-10) when, scoring is 75 points, total word frequency between [10- ∞) when, scoring is 100 points；

Judgment module judgment rule is all in patent application index item: total word frequency is that 0 news commentary is divided into 0 point, total word frequency between When (0-5), scoring is 10 points, total word frequency between [5-10) when, scoring is 25 points, total word frequency between [10-15) when, scoring is 50 Point, total word frequency between [15-20) when, scoring is 75 points, total word frequency between [20- ∞) when, scoring is 100 points.

In one preferred embodiment, enterprise development prospect scoring submodule 33 includes enterprise's investment and financing index dimension Judging part 331, product renewing iteration index dimension judging part 332, product life cycle index dimension judging part 333 and capital city Field dynamic indicator dimension judging part 334.The subclass weight of enterprise's investment and financing index dimension is 0.25, and the product renewing changes Subclass weight for index dimension is 0.25, and the subclass weight of the product life cycle index dimension is 0.25, the capital The subclass weight of market trend index dimension is 0.25.

Wherein, enterprise's investment and financing index dimension judging part 331 includes:

Investments abroad index item, index keyword therein include registering capital to, investing；

Corporate finance index item, index keyword therein include Public Listing, IPO, issue shares, issue bond, day Make wheel, A wheel, B wheel, C wheel, D wheel；

Product renewing iteration index dimension judging part 332 includes:

New technology index item, index keyword therein include new technology investment, new technology, technological change, technological revolution；

Industry barrier breaks through index item, and index keyword therein includes breaking industrial barrier, breaking through barrier；

New product index item, index keyword therein include product news conference；

Product life cycle index dimension judging part 333 includes:

Input time index item, index keyword therein include pouring money, burning money, put goods on the market；

Maturity period index item, index keyword therein include share price rise sharply, price competition, repurchase rate；

Decline phase index item, index keyword therein include that sales volume is decreased obviously；

Capital market dynamic indicator dimension judging part 334 includes:

Positive dynamic indicator item, index keyword therein includes limit-up, market value skyrockets, share price rises violently, finances；

Negative dynamic indicator item, index keyword therein include suspension, merger, recombination, ups and downs, achievement downslide, profit It glides, sale decline, market value is shunk, low-priced valence is sold oneself, diving, continuous drop, in debt, debt promise breaking greatly；

Market conditions good index item, index keyword therein include " liter ", " rising ",

The bad index item of market conditions, index keyword therein include " falling ".

Wherein it is preferred to which judgment module judgment rule is all in investments abroad index item and Corporate finance index item: total word Frequency be 0 news commentary be divided into 0 point, when total word frequency is between (0-3), scoring is 10 points, total word frequency between [3-5) when, scoring is 25 points, always Word frequency between [5-7) when, scoring is 50 points, total word frequency between [7-10) when, scoring is 75 points, total word frequency between [10- ∞) when, Scoring is 100 points；

New technology index item, industry barrier break through judgment module judgment rule in index item and new product index item: Total word frequency is that 0 news commentary is divided into 0 point, when total word frequency is between (0-3), and scoring is 30 points, total word frequency between [3-5) when, scoring is 50 Point, total word frequency between [5- ∞) when, scoring is 100 points；

Judgment module judgment rule is all in input time index item: total word frequency is that 0 news commentary is divided into 0 point, and total word frequency is between (0- 3) when, scoring is 30 points, total word frequency between [3- ∞) when, scoring is 50 points；

Judgment module judgment rule is all in maturity period index item: total word frequency is that 0 news commentary is divided into 0 point, and total word frequency is between (0- 3) when, scoring is 50 points, total word frequency between [3- ∞) when, scoring is 100 points；

Judgment module judgment rule is all in decline phase index item: total word frequency is that 0 news commentary is divided into 0 point, and total word frequency is between (0- 3) when, scoring is 10 points, total word frequency between [3- ∞) when, scoring is 25 points；

Judgment module judgment rule is all in positive dynamic indicator item and market conditions good index item: when total word frequency is 0 Scoring is 0 point, and when total word frequency is between (0-5), scoring is 10 points, total word frequency between [5-10) when, scoring is 25 points, and total word frequency is situated between In [10-15) when, scoring is 50 points, total word frequency between [15-20) when, scoring is 75 points, total word frequency between [20- ∞) when, comment It is divided into 100 points；

Judgment module judgment rule is all in negative dynamic indicator item and the bad index item of market conditions: when total word frequency is 0 Scoring be 0 point, when total word frequency is between (0-5), scoring is -10 points, total word frequency between [5-10) when, scoring is -25 points, total word frequency Between [10-15) when, scoring is -50 points, total word frequency between [15-20) when, scoring is -75 points, total word frequency between [20- ∞) When, scoring is -100 points.

In one preferred embodiment, industry development environment scoring submodule 34 includes that industry prospect index dimension is sentenced Disconnected portion 341 and national policy index dimension judging part 342；The subclass weight of the industry prospect index dimension is 05, the state The subclass weight of family's policy index dimension is 0.5.

Wherein, the industry prospect index dimension judging part 341 includes:

Industry index item, index keyword therein include have a extensive future, unclear prospect it is bright, can also extend including Have very promising prospects, prospect can phase, promise well, and be functionally identical to have a extensive future；

Industry analysis index item, index keyword therein include quickly emerge, development is steady, develop slowly, industry by Resistance can also extend and hold in both hands including emergence, transition and upgrade, growth, Fast Growth, outburst, heat, and be functionally identical to quickly emerge；

The national policy index dimension judging part 342 includes:

Support policy index item, index keyword therein include finance supporting, nursery finance, deduction and exemption enterprise income tax, Exempt enterprise income tax；

Restrictive policy index item, index keyword therein include policies and regulations limitation, policy limitation, restrictive policy；

Protective policy index item, index keyword therein include protective policy, protection in policy；

Adjustment policy index item, index keyword therein include adjustment policy, policy adjustment；

Policy for promotion index item, index keyword therein include policy promotion, policy for promotion；

Guide policy index item, index keyword therein include guide policy, policy guide.

Wherein it is preferred to which judgment module judgment rule is in Industry index item: the word that index keyword has a extensive future Frequency between [1- ∞) news commentary is divided into 100 points；The word frequency that index keyword has a extensive future is 0, and index keyword unclear prospect is bright Word frequency between [1- ∞) news commentary is divided into 50 points；Index keyword has a extensive future and unclear prospect is bright when to be all word frequency be 0, comments It is divided into 0 point；

Judgment module judgment rule is in industry analysis index item: the word frequency that index keyword quickly emerges between [1- ∞) The news commentary is divided into 100 points；The word frequency that index keyword quickly emerges is 0, and index keyword develops stable word frequency between [1- ∞) news commentary is divided into 75 points；The index keyword stable word frequency that quickly emerges and develop is 0, and index keyword develops slowly Word frequency between [1- ∞) news commentary is divided into 50 points；It is 0 that index keyword, which quickly emerges, develops word frequency that is steady and developing slowly, and The word frequency that index keyword industry is obstructed between [1- ∞) news commentary is divided into 25 points；Total word frequency is that 0 news commentary is divided into 0 point；

Mould is judged in support policy index item, adjustment policy index item, policy for promotion index item and guide policy index item Block judgment rule is: total word frequency between [1- ∞) news commentary is divided into 100 points；

Restrictive policy index item, middle judgment module judgment rule is: total word frequency between [1- ∞) news commentary is divided into 0 point.

The specific indexes keyword and judgment module judgment rule limited according to the present invention, final score 50 divide it is above i.e. It can be referred to as qualification, if being higher than 60 points can be referred to as outstanding, in fact Some Enterprises scoring can be negative point.

It further include computing module and display device in the system, wherein the computing module and each described scoring submodule Block is connected, the scoring provided to obtain each scoring submodule, and according to the major class weight coefficient of each scoring submodule Final overall score is calculated, the display device is to show input information, such as enterprise name, also to show participle number Amount, the index keyword quantity of hit and final overall score etc. information.

A kind of business risk evaluation method based on open source data is also provided in the present invention, this method is by described above What the business risk evaluation system based on open source data was realized.

Embodiment:

By taking PetroChina Company Ltd. as an example, business risk assessment is carried out, it is public that data crawl module input Title PetroChina Company Ltd./China Petroleum is taken charge of, associated nets number of pages is 198, crawls article totally 3574908 , keyword quantity: 104477, final score: 77.75；

Wherein, highest 100 keywords of word frequency are as follows: (China, 1704), (enterprise, 967), (petroleum, 654), (company, 621), (center, 510), (natural gas, 504), (development, 504), (country, 498), (making an inspection tour, 423), (price, 419) (changes Leather, 394), (market, 380), (work, 373), (construction, 352), (group company, 333), (group, 333), (it is economical, 312), (secretary, 303), (problem, 298), (reporter, 287), (cooperation, 277), (Co., Ltd, 266), (project, 265), (energy, 255), (crude oil, 243), (industry, 239), (international, 237), (carrying out, 230), (indicating, 227), (general manager, 224), (currently, 224), (realizing, 221), (state-owned, 221), (this year, 221), (central enterprise, 216), (technology, 212), (it is important, 211), (president, 211), (China, 209), (management, 204), (situation, 201), (industry, 199), (production capacity, 191), (party Group, 186), (passing through, 183), (one, 177), (first, 177), (2015,176), (meanwhile 175), (resource, 175), (Beijing, 174), (whole nation, 174), (field, 168), (middle petroleum, 167), (promoting, 166), (personnel, 166), (it leaves office, 166), (all the way, 165), (university, 165), (state-owned enterprise, 165), (oil gas, 162), (pipeline, 161), (wherein, 159), (investment, 158), (exploitation, 156), (lease, 155), (oil field, 152), (business, 151), (Iran, 150), (becoming, 150), (discipline inspection commission, 149), (area, 149), (increasing, 147), (thinking, 147), (special, 144), (product, 144), (oil price, 144), (production, 141), (leader, 140), (aspect, 140), (capital, 139), (party committee, 139), (mechanism, 139), (, 136), (dollar, 136), (center, 135), (petrochemical industry, 131), (society, 130), (service, 129), (unit, 128), (providing, 128), (department, 127), (as 125), (since, 125), (main, 124), (responsibility, 123), (research, 122), (structure, 120), (it is horizontal, 120), (adjustment, 120).

Each index dimension scores:

Enterprise's key person: 30.0

Corporate social reputation: 13.5

Public records: -76.0

Enterprise innovation is horizontal: 55.0

Brand influence: 55.0

Enterprise's investment and financing: 12.5

Product renewing iteration: 0.0

Product life cycle: 0.0

Capital market dynamic: 0.0

Industry prospect: 25.0

National policy: 200.0；

Wherein, each index item score:

Dong supervises high internet exposure: 100=summit: 5+ forum: 26+ meeting: 3+ innovates conference: 0+ special visit: 5+ product News conference: 0+ seminar: 6+ development conference: 0

Social responsibility: 100=public good: 10+ is charitable: 1+ industry leader: 0+ model worker: 1

Regional telephone distribution: -100=leaves office: 166+ resigns: 0+ is negative: 2+ disturbance: 3+ fails: 1+ is undisciplined: 0+ is looked into: 0 + investigated: 1

Positive news information: 0=Outstanding Contribution Award: 0+ man of the hour: 0+ outstanding person: 0+ leader: 0+ is most preferably high Pipe: 0+ annual character: 0

Winning information: 0=wins a prize enterprise: 0+ silver medal: 0+ gold medal: 0+ Outstanding Contribution Award: 0+ special award Innovation Awards: 0+ is outstanding Manage prize: 0+ medal: 0+ honorary certificate: 0+ prize-giving grand ceremony: 0

Commend information: 10=medal: 0+ honorary certificate: 0+ favorable comment: 1

Reputation reputation: 25=is well received: 0+ favorable comment: 1+ phenomenon grade: 0+ public praise prize: 0+ is best: 5+ enterprise public praise: 0+ net Network public praise: 0

Commonweal information: 10=public good Contribution Prize: 0+ is charitable: 1+ donations: 0+ donation: 0+ public welfare activities: 0+ utility: 0

Business ethics: 0=business ethics enterprise: 0+ morals enterprise: 0+ business ethics is ruined: 0

Business ethics is ruined: 0=is ruined: 0

Administrative penalty information: 0=administrative penalty: 0+ responsibility dispute: 0+ illegal operation: 0+ is accused of in violation of rules and regulations: 0+ punishes publicity: 0

Administrative permission information: 10=operation permission: 0+ administrative permission: 0+ licensing: 1

Manage exception information: 0=manages abnormal register: 0+ manages register extremely: 0+ exception distributors: 0+ manages abnormal Enterprise: 0

Tax negative information: 0=pays taxes exception: 0+ tax evasion: 0+ tax declaration is abnormal: 0

Media negative information: -100=is accused of: 94+ substandard product: 0+ is recalled: 0+ rectification: 45+ labour dispute: 0+ is negative Face news: 0+ reduces the staff: 3+ runs away: 0+ fakes: 0

Persecutio information: -100=infringement: 1+ loses a lawsuit: 0+ prosecution: 22+ lawsuit: 9

Patent application: 10=patent: 2+ patent of invention: 0+ patent certificate: 0

Trade mark registration: 0=trade mark: 0+ trademark application: 0

Copyright is delivered: 100=copyright: 11+ copyright: 0

Brand recognition: 10=popularity: 2+ esbablished corporation: 0+ well-known trademark: 0+ freed-from-inspection product: 0+ reputation: 0

Brand share: 100=occupation rate of market: 4+ monopolization: 33

Investments abroad: 0=is registered capital to: 0+ investment monopolization: 0

Corporate finance: 50=Public Listing: 0+IPO:5+ floating stocks: 0+ issues bond: 0+ angel wheel: 0+A wheel: 0+B Wheel: 0+C wheel: 0+D wheel: 0

New technology: 0=new technology investment: 0+ new technology: 0+ technological change: 0+ technological revolution: 0

Industry barrier is broken through: 0=breaks industrial barrier: 0+ breakthrough barrier: 0

New product: 0=product news conference: 0

Input time: 0=pours money: 0+ burns money: 0+ puts goods on the market: 0

Maturity period: 0=share price rises sharply: 0+ price competition: 0+ repurchase rate: 0

Decline phase: 0=sales volume is decreased obviously: 0

Positive dynamic: 100=limit-up: 3+ market value skyrockets: 0+ share price rises violently: 0+ financing: 63

Negative dynamic: -100=is suspended: 3+ is merged: 15+ recombination: 63+ ups and downs: 0+ achievement glides: 0+ declination of profits: 0+ pin Sell decline: 0+ market value is shunk: 0+ is low-priced, and valence is sold oneself: the big diving of 0+: 0+ continuously drops: 0+ is in debt: 0+ debt promise breaking: 0

Market conditions are good: 0=liter: 0+ rises: 0

Market conditions are bad: 0=falls: 0

Industry: 50=has a extensive future: 0+ unclear prospect is bright: 1

Industry analysis: 0=quickly emerges: 0+ development is steady: 0+ is developed slowly: 0+ industry is obstructed: 0

Support policy: 100=finance supporting: 17+ nursery finance: 3+ reduces or remits enterprise income tax: 0+ exempts enterprise income tax: 0

Restrictive policy: -100=policies and regulations limitation: 10+ policy limitation: 17

Protective policy: 100=protective policy: 8+ protection in policy: 10

Adjustment policy: 100=adjusts policy: 2+ policy adjustment: 1

Policy for promotion: 100=policy promotes: 1+ policy for promotion: 7

Guide policy: 100=guide policy: 2+ policy guide: 1

Questionnaire survey is done to the employee in PetroChina Company Ltd., specific post, inside again, including method Business department employee, Finance Department employee, administrative department employee, personnel department employee, business department employee, middle level manager and portion Divide total 100 people such as branch office representative, include the content in each index item of the present invention in specific questionnaire table, is i.e. Dong supervises height Internet exposure index item, regional telephone distribution index item, positive news information index item, obtains social responsibility index item It encourages information index item, commend information index item, reputation reputation index item, commonweal information index item, business ethics index item, administration It punishes information index item, administrative permission information index item, manage exception information index item, tax negative information index item, media Negative information index item, persecutio information index item, patent application index item, trade mark registration index item, copyright deliver finger Mark item, brand recognition index item, brand share index item, investments abroad index item, Corporate finance index item, new technology refer to Mark item, industry barrier break through index item, new product index item, input time index item, maturity period index item, decline phase index item, Positive dynamic indicator item, negative dynamic indicator item, market conditions good index item, the bad index item of market conditions, Industry Index item, industry analysis index item, support policy index item, restrictive policy index item, protective policy index item, adjustment policy refer to Mark item, policy for promotion index item, guide policy index item；

There are 5 options selective in each index item, in social responsibility index item, including social responsibility is very Height, social responsibility is higher, social responsibility is general, social responsibility is not high and does not know；For another example in negative dynamic indicator item, Including negatively dynamic, very much, negative dynamically more, negative dynamic is generally more, negative dynamic is few and does not know；

After taking the questionnaire filled, count the number selected in each index item, except " not knowing " option with Outside, the option for selecting number most is existed as the index item questionnaire final result if there is the identical situation of number with position Final result of the preceding option as the index item；Final statistics is obtained such as following table one:

Table one

Four options respectively correspond 100 points, 65 points, 30 points and 0 point after wherein score value causes for the past of the index item of positive value, Score value is that four options respectively correspond -100 points, -65 points, -30 points and 0 point after causing the past of the index item of negative value, according still further to this Group weight and major class weight calculation in invention obtain final score, and the final score of above-mentioned questionnaire survey is in the present invention 82.3 points；

The reliability and reasonability of system and method are provided in order to further verify the present invention, also other more companies are used System provided by the invention has done risk assessment, and does questionnaire survey respectively to the employee inside more companies, obtain as Comparing result shown in following table two；

Table two

Business Name	System evaluation score	Questionnaire survey score
			China PetroChemical Corporation	77.75	82.3
State Grid Corporation of China	80.16	88.1
			China Mobile communicates group company	75.83	84.25
Chinese railway construction parent company	79.51	85.87
			China life insurance (group) company	76.1	80.32
Beijing automobile group	73.5	79.2
			China Datang Power Group Corporation	74.19	78.85

The all relatively low 4-8 of questionnaire survey points according to result assessment system score provided by the invention relative to employee Left and right, but overall score fluctuation is more stable, and each good company's score of management state is not much different, and can illustrate the present invention The system of offer have high reasonability and stability, from the point of view of above-described embodiment, according to system provided by the invention into Row assessment, if score, which is higher than 70 points, is believed that the situation of enterprises is good, 75 points or more are regarded as superior level.

Methods of marking provided by the invention and the Questionnaire results are further analyzed it is found that the major class weight coefficient There is great influence for final appraisal result with each group weight coefficient, for example, enterprise operation and management and industry development The importance of environment is all higher than the importance of enterprise competitiveness and development prospect, in finishing analysis mass data and combines Major class weight coefficient of the invention is designed in the case where China's actual conditions, to pass through weight distribution proportional balancing method various aspects Relatively important relationship, i.e., to the influence degree of business risk；

In each grading module, corresponding group weight coefficient is set for different submodule respectively, be equally for The percentage contribution influenced between each scoring submodule of balance for business risk, wherein score son in enterprise operation and management In module, the public records of enterprise more can scientific, objectively embody the warp of enterprise relative to enterprise key person and social reputation Battalion's situation, the influence for business risk are bigger；In addition, it is contemplated that the source of data, analyzes data according to data source The characteristics of, and then select, the reasonable judgement of setting, code of points, such as judgment module is sentenced in setting national policy index item When disconnected rule, it is to be understood that bigger to the difficulty for crawling the acquisition policy information by network, the information content of acquisition is less, accordingly Interference information also can be less, so have a small amount of keyword hit when can provide higher score value；In addition, for positive new Hear judgment module judgment rule in the projects such as information index item be all hit keyword reach sufficient amount Shi Caineng provide compared with High score value, and multiple score value gears are set, it is more scientific reasonable finally to score, reduce the shadow of interference information It rings；

So business risk evaluation system provided by the invention and method are for analyzing influence enterprise wind more fully hereinafter The factor of danger, is more quickly obtained business risk evaluation result, the more specific gravity between reasonable distribution business risk influence factor Relationship, and then obtain business risk evaluation result scientific and reasonable, close to truth；

On the basis of mass data analysis, in conjunction with actual conditions, it is provided in system and method provided by the invention The multiple grading module, multiple scoring submodules and corresponding judgment module and judgment rule in the present invention；

System and method in the finally obtained present invention can obtain comprehensively and enterprise on the basis of convenient and efficient Relevant data information, and scientific and reasonable weight distribution is made to data information, and then acquisition is more valuable, is more nearly The business risk evaluation result of truth, but also can rapidly other relevant enterprises of lateral comparison, by comparing different The final score value of enterprise can understand risk size opposite between each enterprise cheer and brightly.

Combining preferred embodiment above, the present invention is described, but these embodiments are only exemplary , only play the role of illustrative.On this basis, a variety of replacements and improvement can be carried out to the present invention, these each fall within this In the protection scope of invention.

Claims

1. a kind of business risk evaluation system based on open source data, which is characterized in that the system includes

Data crawl module (1), are used to crawl data from webpage,

Data word segmentation module (2) is used to crawl data the data text that module (1) crawls and does word segmentation processing, and counts Word frequency；With

Grading module (3) is used to provide enterprise's scoring according to the word frequency of participle.

2. the business risk evaluation system according to claim 1 based on open source data, which is characterized in that

It is crawled in the data and is circumscribed with input unit (11) on module (1), input retrieval letter by the input unit (11) Breath.

3. the business risk evaluation system according to claim 1 based on open source data, which is characterized in that

Institute's scoring module (3) include enterprise operation and management scoring submodule (31), enterprise competitiveness scoring submodule (32), Enterprise development prospect scores submodule (33) and industry development environment scores one or more of submodule (34), respectively obtains The scoring of each submodule included by grading module (3) sums up to obtain the most final review of the enterprise according still further to major class weight Point；

Wherein it is preferred to the major class weight coefficient of enterprise operation and management scoring submodule (31) is 0.4, competition among enterprises energy The major class weight coefficient of power scoring submodule (32) is 0.2, the major class weight coefficient of enterprise development prospect scoring submodule (33) It is 0.1, the major class weight coefficient of industry development environment scoring submodule (34) is 0.3.

4. the business risk evaluation system according to claim 3 based on open source data, which is characterized in that

It all include more than two index dimension judging parts in each submodule,

All include more than two index item in each index dimension judging part, and is scored respectively each index item；

Subclass multiplied by weight after the scoring of each index item in one index dimension judging part is added with the index dimension, obtains The scoring of the index dimension；

The sum of each index dimension scoring for the submodule scoring.

5. the business risk evaluation system according to claim 4 based on open source data, which is characterized in that

More than one index keyword, and point extracted in data word segmentation module (2) are all stored in each index item Participle identical with index keyword is found out in word, and knows the word frequency of the participle；

Preferably, judgment module is also stored in each index item, the judgment module is according to total word frequency or hit The content of keyword judges the scoring of each index item；

Total word frequency is the sum of the word frequency of all index keyword correspondence/hit participles in index item.

6. the business risk evaluation system according to claim 5 based on open source data, which is characterized in that

Enterprise operation and management scoring submodule (31) includes enterprise's key person index dimension judging part (311), corporate social reputation Index dimension judging part (312) and public records index dimension judging part (313)；

Wherein, enterprise's key person index dimension judging part (311) includes:

Dong supervises high internet exposure index item, and index keyword therein includes summit, forum, annual meeting, innovates conference, is special Visit, seminar, develops conference at product news conference,

Social responsibility index item, index keyword therein include public good, charitable, industry leader, model worker,

Regional telephone distribution index item, index keyword therein include leaving office, resigning, is negative, disturbance, fail, is undisciplined, quilt It looks into, investigated, and

Positive news information index item, index keyword therein include Outstanding Contribution Award, man of the hour, outstanding person, leader Personage, best senior executive, annual character；

The corporate social reputation index dimension judging part (312) includes:

Winning information index item, index keyword therein include prize-winning enterprise, silver medal, gold medal, Outstanding Contribution Award, special award wound New prize, outstanding operation prize, medal, honorary certificate, prize-giving grand ceremony,

Information index item is commended, index keyword therein includes medal, honorary certificate, favorable comment,

Reputation reputation index item, index keyword therein include well received, favorable comment, phenomenon grade, public praise prize, best, enterprise Public praise, internet word-of-mouth,

Commonweal information index item, index keyword therein include public good Contribution Prize, charitable, donations, donation, public welfare activities, public affairs Beneficial cause,

Business ethics ruins index item, and index keyword therein includes that business ethics is ruined, ruined；

The public records index dimension judging part (313) includes:

Administrative penalty information index item, index keyword therein include administrative penalty, responsibility dispute, illegal operation, are accused of disobeying Rule, punishment publicity,

Administrative permission information index item, index keyword therein include operation permission, administrative permission, licensing,

Exception information index item is managed, index keyword therein manages register, abnormal operation including the abnormal register of operation, exception Enterprise manages abnormal enterprise,

Tax negative information index item, index keyword therein is abnormal including exception of paying taxes, tax evasion, tax declaration,

Media negative information index item, index keyword therein include being accused of, substandard product, recalling, rectify and improve, working and entangle Confusingly, negative press, reduce the staff, run away, faking, and

Persecutio information index item, index keyword therein include encroach right, lose a lawsuit, prosecuting, lawsuit；

Wherein it is preferred to

Dong supervises judgment module judgment rule in high internet exposure index item and social responsibility index item: total word frequency is 0 news commentary is divided into 0 point, when total word frequency is between (0-3), and scoring is 30 points, total word frequency between [3-5) when, scoring is 50 points, total word frequency Between [5- ∞) when, scoring is 100 points；

Judgment module judgment rule is all in winning information index item, commendation information index item and commonweal information index item: total word Frequency be 0 news commentary be divided into 0 point, when total word frequency is between (0-3), scoring is 10 points, total word frequency between [3-5) when, scoring is 25 points, always Word frequency between [5-7) when, scoring is 50 points, total word frequency between [7-10) when, scoring is 75 points, total word frequency between [10- ∞) when, Scoring is 100 points；

Positive news information index item, business ethics good index item, reputation reputation index item and administrative permission information index item Middle judgment module judgment rule is all: total word frequency is that 0 news commentary is divided into 0 point, when total word frequency is between (0-5), and scoring is 10 points, total word Frequency between [5-10) when, scoring is 25 points, total word frequency between [10-15) when, scoring is 50 points, total word frequency between [15-20) when, Scoring be 75 points, total word frequency between [20- ∞) when, scoring is 100 points；

Regional telephone distribution index item, business ethics ruin index item, administrative penalty information index item, manage exception information index Judgment module judgement rule in item, tax negative information index item, media negative information index item and persecutio information index item All be then: total word frequency is that 0 news commentary is divided into 0 point, when total word frequency is between (0-5), and scoring is -10 points, total word frequency between [5-10) when, Scoring be -25 points, total word frequency between [10-15) when, scoring is -50 points, total word frequency between [15-20) when, scoring is -75 points, Total word frequency between [20- ∞) when, scoring is -100 points.

7. the business risk evaluation system according to claim 5 based on open source data, which is characterized in that

Enterprise competitiveness scoring submodule (32) includes the horizontal index dimension judging part (321) of enterprise innovation and brand influence Index dimension judging part (322)；

Wherein, the horizontal index dimension judging part (321) of the enterprise innovation includes:

Patent application index item, index keyword therein include patent, patent of invention, patent certificate,

Trade mark registration index item, index keyword therein include trade mark, trademark application,

Copyright delivers index item, and index keyword therein includes copyright, copyright；

The brand influence index dimension judging part (322) includes:

Brand recognition index item, index keyword therein include popularity, esbablished corporation, well-known trademark, freed-from-inspection product, beauty Reputation degree,

Brand share index item, index keyword therein include occupation rate of market, monopolization；

Wherein it is preferred to

Trade mark registration index item, copyright are delivered to be judged in index item, brand recognition index item and brand share index item Module judgment rule is all: total word frequency is that 0 news commentary is divided into 0 point, when total word frequency is between (0-3), and scoring is 10 points, total word frequency between [3-5) when, scoring is 25 points, total word frequency between [5-7) when, scoring is 50 points, total word frequency between [7-10) when, scoring is 75 Point, total word frequency between [10- ∞) when, scoring is 100 points；

Judgment module judgment rule is all in patent application index item: total word frequency is that 0 news commentary is divided into 0 point, and total word frequency is between (0-5) When, scoring is 10 points, total word frequency between [5-10) when, scoring is 25 points, total word frequency between [10-15) when, scoring is 50 points, always Word frequency between [15-20) when, scoring is 75 points, total word frequency between [20- ∞) when, scoring is 100 points.

8. the business risk evaluation system according to claim 5 based on open source data, which is characterized in that

Enterprise development prospect scoring submodule (33) includes enterprise's investment and financing index dimension judging part (331), product renewing iteration Index dimension judging part (332), product life cycle index dimension judging part (333) and the judgement of capital market dynamic indicator dimension Portion (334)；

Wherein, enterprise investment and financing index dimension judging part (331) include:

Corporate finance index item, index keyword therein include Public Listing, IPO, issue shares, issue bond, angel take turns, A wheel, B wheel, C wheel, D wheel；

Product renewing iteration index dimension judging part (332) includes:

Product life cycle index dimension judging part (333) includes:

Capital market dynamic indicator dimension judging part (334) includes:

Positive dynamic indicator item, index keyword therein includes limit-up, market value skyrockets, share price rises violently, finances,

Negative dynamic indicator item, index keyword therein include suspension, merger, recombination, ups and downs, achievement downslide, declination of profits, Sale decline, market value are shunk, low-priced valence is sold oneself, diving, continuous drop, in debt, debt promise breaking greatly,

The bad index item of market conditions, index keyword therein include " falling "；

Wherein it is preferred to

Judgment module judgment rule is all in investments abroad index item and Corporate finance index item: total word frequency is that 0 news commentary is divided into 0 Point, when total word frequency is between (0-3), scoring is 10 points, total word frequency between [3-5) when, scoring is 25 points, total word frequency between [5-7) When, scoring is 50 points, total word frequency between [7-10) when, scoring is 75 points, total word frequency between [10- ∞) when, scoring is 100 points；

New technology index item, industry barrier break through judgment module judgment rule in index item and new product index item: total word Frequency be 0 news commentary be divided into 0 point, when total word frequency is between (0-3), scoring is 30 points, total word frequency between [3-5) when, scoring is 50 points, always Word frequency between [5- ∞) when, scoring is 100 points；

Judgment module judgment rule is all in input time index item: total word frequency is that 0 news commentary is divided into 0 point, and total word frequency is between (0-3) When, scoring is 30 points, total word frequency between [3- ∞) when, scoring is 50 points；

Judgment module judgment rule is all in maturity period index item: total word frequency is that 0 news commentary is divided into 0 point, and total word frequency is between (0-3) When, scoring is 50 points, total word frequency between [3- ∞) when, scoring is 100 points；

Judgment module judgment rule is all in decline phase index item: total word frequency is that 0 news commentary is divided into 0 point, and total word frequency is between (0-3) When, scoring is 10 points, total word frequency between [3- ∞) when, scoring is 25 points；

Judgment module judgment rule is all in positive dynamic indicator item and market conditions good index item: total word frequency scores when being 0 Be 0 point, when total word frequency is between (0-5), scoring is 10 points, total word frequency between [5-10) when, scoring is 25 points, total word frequency between [10-15) when, scoring is 50 points, total word frequency between [15-20) when, scoring is 75 points, total word frequency between [20- ∞) when, scoring It is 100 points；

Judgment module judgment rule is all in negative dynamic indicator item and the bad index item of market conditions: total word frequency scores when being 0 Be 0 point, when total word frequency is between (0-5), scoring is -10 points, total word frequency between [5-10) when, scoring is -25 points, total word frequency between [10-15) when, scoring is -50 points, total word frequency between [15-20) when, scoring is -75 points, total word frequency between [20- ∞) when, comment It is divided into -100 points.

9. the business risk evaluation system according to claim 5 based on open source data, which is characterized in that

Industry development environment scoring submodule (34) includes industry prospect index dimension judging part (341) and national policy index dimension It spends judging part (342)；

Wherein, the industry prospect index dimension judging part (341) includes:

Industry index item, index keyword therein is including having a extensive future, unclear prospect is bright；

Industry analysis index item, index keyword therein includes quickly emergence, development is steady, develop slowly, industry is obstructed；

The national policy index dimension judging part (342) includes:

Support policy index item, index keyword therein include finance supporting, nursery finance, deduction and exemption enterprise income tax, exempt Enterprise income tax,

Restrictive policy index item, index keyword therein include that policies and regulations limit, policy limits, restrictive policy,

Protective policy index item, index keyword therein include protective policy, protection in policy,

Adjustment policy index item, index keyword therein include adjustment policy, policy adjustment,

Policy for promotion index item, index keyword therein include policy promote, policy for promotion, and

Guide policy index item, index keyword therein include guide policy, policy guide；

Wherein it is preferred to

Judgment module judgment rule is in Industry index item: the word frequency that index keyword has a extensive future between [1- ∞) news commentary It is divided into 100 points；The word frequency that index keyword has a extensive future is 0, and the bright word frequency of index keyword unclear prospect between [1- ∞) The news commentary is divided into 50 points；Index keyword has a extensive future and unclear prospect is bright when to be all word frequency be 0, and scoring is 0 point；

Judgment module judgment rule is in industry analysis index item: the word frequency that index keyword quickly emerges between [1- ∞) news commentary It is divided into 100 points；The word frequency that index keyword quickly emerges is 0, and index keyword develop stable word frequency between [1- ∞) when Scoring is 75 points；The index keyword stable word frequency that quickly emerges and develop is 0, and the word frequency that develops slowly of index keyword is situated between In [1- ∞) news commentary is divided into 50 points；It is 0 that index keyword, which quickly emerges, develops word frequency that is steady and developing slowly, and index is closed The word frequency that key word industry is obstructed between [1- ∞) news commentary is divided into 25 points；Total word frequency is that 0 news commentary is divided into 0 point；

Judgment module is sentenced in support policy index item, adjustment policy index item, policy for promotion index item and guide policy index item Disconnected rule is: total word frequency between [1- ∞) news commentary is divided into 100 points；

10. a kind of business risk evaluation method based on open source data, this method is by any one of such as claim 1-9 What the business risk evaluation system based on open source data was realized.