CN104820629A - Intelligent system and method for emergently processing public sentiment emergency - Google Patents
Intelligent system and method for emergently processing public sentiment emergency Download PDFInfo
- Publication number
- CN104820629A CN104820629A CN201510243751.0A CN201510243751A CN104820629A CN 104820629 A CN104820629 A CN 104820629A CN 201510243751 A CN201510243751 A CN 201510243751A CN 104820629 A CN104820629 A CN 104820629A
- Authority
- CN
- China
- Prior art keywords
- public sentiment
- text
- classification
- network text
- feature
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Abstract
The invention discloses an intelligent system and an intelligent method for emergently processing a public sentiment emergency, which applies natural language processing technology, ontological theory and semantic association technique to intelligent identification of the internet public sentiment emergency and automatic generation of a preventing and controlling scheme, wherein the system and the method of the invention can realize formatting conversion of the preventing and controlling scheme for the emergency based on a computer information processing method, realize the semantic matching between the circumstance of the public sentiment emergency and scheme, realize the accurate recognition of all kinds of internet public sentiment emergency, and aid to making a decision. The system and the method of the invention can be used for monitoring the internet public sentiment in real time, aiding to making a preventing and controlling decision for the internet public sentiment, and improving the preventing and controlling response speed of the public sentiment emergency.
Description
Technical field
The invention belongs to computer application field, relate to and natural language processing technique, ontology theory and semantic association technology are applied to the Intelligent Recognition of internet public feelings accident and the automatic generation of prevention and control scheme.It realizes transforming the format of emergent prevention and control prediction scheme based on computer information processing method, realizes the semantic matches between public sentiment accident sight and prediction scheme, realizes the accurate identification to various internet public feelings accident and aid decision making.
Background technology
Along with the development of Internet technology, internet has become a kind of mass media be widely used, and its feeler almost stretches to the every field of society, and the important medium that becomes public opinion gradually new.Network public-opinion be the public on the internet open express to certain social phenomenon or social concern, there is certain influence power and tendentious communis opinio, the impact of network public-opinion on political life order and social stability grows with each passing day, some network public-opinion accidents can not be dealt carefully with in time, very likely bring out the unhealthy emotion of the common people and the generation of bad behavior, and then serious threat is formed to social stability.The automatic monitoring to network public sentiment information can be realized in the urgent need to a kind of technological means, decision support can be provided to the disposal of public sentiment accident.
Summary of the invention
The present invention is exactly for the demand, propose a kind of computer application system-public sentiment accident emergent treatment system, it can be monitored in real time to internet public feelings, aid decision making person can form matched prevention and control scheme targetedly according to the actual conditions of public sentiment accident, accelerate the disposal response speed to network public-opinion accident.
Technical matters to be solved by this invention is realized by following technical scheme:
A public sentiment accident emergent treatment system for intelligence, is characterized in that: this system comprises internet information acquisition and parsing module, internet information analysis module, network text classification judge and Cluster Analysis module, emergence treatment scheme generation module and emergency processing recruitment evaluation module; Described internet information acquisition and parsing module, for from Information Monitoring on internet, extract the metadata information of natural language word and webpage in webpage, and are saved in database; The natural language word that described internet information analysis module is used for gathering in next information carries out feature extraction, forms text feature; Described network text classification judges to be used for judging the classification of network text with Cluster Analysis module, carries out cluster analysis to cumulative network text; Described emergence treatment scheme generation module is used for automatically generating according to the concrete condition of public sentiment event processing prediction scheme accordingly, and decision-maker can based on process prediction scheme formulation and implementation scheme; Described emergency processing recruitment evaluation module is used for assessing the implementation effect carried into execution a plan.
The public sentiment accident emergent treatment system of intelligence and a method, is characterized in that the method comprises the following steps:
1. internet information acquisition and parsing: gather the network datas such as forum postings, Blog content and Website News webpage from internet forum, blog, news website by the computing machine of connecting Internet, then, computing machine is utilized to adopt rule-based information extraction technique automatically to resolve network data, from wherein extracting two category informations: the metadata information of natural language Word message and webpage; Natural language Word message comprises the information such as headline, body, forum postings title, model content; The metadata information of webpage comprises the information such as web site name, website URL of the time of delivering, author, posting person, model reply volume, model amount of reading, appearance, the information parsed is saved in database, information acquisition and parsing are lasting processes, are formed and monitor the Automatic continuous of internet site;
2. internet information analysis: first utilize the Chinese word cutting method of natural language processing technique to carry out participle respectively to the title of network text and body matter, and the part of speech of lexical item each in word segmentation result is marked, give up to fall afterwards except noun, verb, lexical item outside adjective, then text multiple-accuracy representing method is utilized to extract the single lexical item characteristic sum lexical item linked character of network text, geographic location feature in network text and character features is identified again according to the part-of-speech tagging situation in word segmentation result, geographic location feature is the geographic position name occurred in network text, character features is the person names occurred in network text,
3. the lexical item in the network text after step 2. being processed is compared with the lexical item feature of the public sentiment classification set in Computer Database and is mated, and according to matching result, network text is carried out classification process according to the public sentiment classification set in Computer Database; The network text that can not sort out carries out cluster analysis, network text close for content is polymerized to bunch, if network amount of text exceeds setting threshold value in bunch, then to bunch in the network text lexical item feature of carrying out public sentiment classification take out process, and the lexical item feature of the public sentiment classification of extraction is added in Computer Database; 4. step is proceeded to for the network text completing classification; Wherein, matching content comprises single lexical item feature, lexical item linked character, geographic location feature and character features;
If 4. at the appointed time in section, belong to the quantity of the network text of a certain classification or occur that the Websites quantity of this classification network text exceedes the threshold value of specifying, then initiate emergency plan;
Complete the emergency processing of intelligent public sentiment accident.
Wherein, after step 4., also comprise emergency processing recruitment evaluation step: first according to evaluation index acquisition index data, then achievement data input assessment formula is drawn quantitative evaluation result.
Wherein, step 3. according to matching result by network text according to Computer Database in the public sentiment classification that sets carry out sorting out process and be specially: the method that network text classification judges the lexical item of network text is compared with the lexical item feature of each public sentiment classification to mate, matching operation is carried out respectively in single word feature, word association feature, geographic location feature and character features four, obtain the Similarity value of network text and each public sentiment classification according to match condition, text is attributed to the public sentiment classification that Similarity value is the highest.
Wherein, step 3. in bunch in the network text lexical item feature of carrying out public sentiment classification take out process, be specially: suppose that the network text that bunch T comprises has T={t
1, t
2... t
n, utilize text multiple-accuracy representing method to extract each text t
isingle lexical item characteristic sum lexical item linked character, statistical method is adopted to calculate the Statistical Distribution of all single lexical item characteristic sum lexical item linked character of all texts in T again, select the vocabulary that occurred in network text over half in T as public sentiment classification lexical item feature, and calculate the frequency of its average occurrence frequency in T as public sentiment category feature lexical item; Wherein, 1≤i≤n.
Wherein, step 4. in the generation method of emergency preplan be: based on internet public feelings event sight ontology knowledge library model and network public-opinion prevention and control measure prediction scheme ontology knowledge base, utilize semantic matches technology according to the concrete condition of public sentiment event sight, from prevention and control measure prediction scheme storehouse, Auto-matching goes out optimal plan for emergency handling.
Compared with prior art, the present invention has following advantage and beneficial effect:
1, the present invention can not only carry out automatic monitoring to network public-opinion, can also provide prevention and control measure scheme for burst public sentiment event.
2, public sentiment type identification Computer Database of the present invention has extensibility, constantly supplements novel public sentiment type feature in database, enable system identify the public sentiment event of newly-increased type by Clustering Analysis of Text.
Accompanying drawing explanation
Fig. 1 system module composition diagram
Fig. 2 public sentiment taxonomic hierarchies illustraton of model
Fig. 3 public sentiment taxonomic hierarchies concept attribute illustraton of model
Fig. 4 public sentiment taxonomic hierarchies schematic diagram
Fig. 5 category feature production process fundamental diagram
Fig. 6 semantic matches schematic diagram
The knowledge augmented figure of Fig. 7 text cluster Network Based
Fig. 8 public sentiment event sight ontology knowledge base figure
Fig. 9 public sentiment prevention and control measure prediction scheme ontology knowledge base figure
Figure 10 network public-opinion prevention and control Knowledge Semantic Model Based figure
Figure 11 is based on the matching process figure of semanteme
Figure 12 emergency processing recruitment evaluation index system figure
Embodiment
Below in conjunction with the drawings and specific embodiments, the present invention will be further described.But embodiments of the present invention are not limited thereto.
The present embodiment provides a kind of public sentiment accident emergent treatment system of intelligence, this system comprises internet information acquisition and parsing module, internet information analysis module, network text classification judges and Cluster Analysis module, emergence treatment scheme generation module, emergency processing recruitment evaluation module, as shown in Figure 1; Described internet information acquisition and parsing module, for from Information Monitoring on internet, extract the metadata information of natural language word and webpage in webpage, and are saved in database; The natural language word that described internet information analysis module is used for gathering in next information carries out feature extraction, forms text feature; Described network text classification judges to be used for judging the classification of network text with Cluster Analysis module, carries out cluster analysis to cumulative network text; Described emergence treatment scheme generation module is used for automatically generating according to the concrete condition of public sentiment event processing prediction scheme accordingly, and decision-maker can based on process prediction scheme formulation and implementation scheme; Described emergency processing recruitment evaluation module is used for assessing the implementation effect carried into execution a plan.
The present embodiment also provides a kind of method of work of public sentiment accident emergent treatment system of intelligence, and the method comprises the following steps:
1. internet information acquisition and parsing: gather the network datas such as forum postings, Blog content and Website News webpage from internet forum, blog, news website by the computing machine of connecting Internet, then, computing machine is utilized to adopt rule-based information extraction technique automatically to resolve network data, from wherein extracting two category informations: the metadata information of natural language Word message and webpage.Natural language Word message comprises the information such as headline, body, forum postings title, model content, author, posting person; The metadata information of webpage comprises the web site name, website URL etc. of the time of delivering, model reply volume, model amount of reading, appearance, the key message parsed is saved in database, information acquisition and parsing are lasting processes, are formed and monitor the Automatic continuous of internet site.
2. internet information analysis: first utilize the Chinese word cutting method of natural language processing technique to carry out participle and part-of-speech tagging process to the title of network text and body matter, mark out the part of speech of each lexical item, give up to fall the vocabulary except noun, verb, adjective in text.Then a kind of " text multiple-accuracy representing method for text retrieval system " that obtained national inventing patent mandate described method is utilized to extract the single word characteristic sum word association feature of network text.In addition, geographic location feature in text and character features is identified according to the part-of-speech tagging situation in word segmentation result, geographic location feature is the geographic position name, the character features that occur in network text is the person names occurred in network text, as shown in the network text semantic feature extract function unit in accompanying drawing 5.Generally speaking the feature of network text is one group of vocabulary, is furnished with its occurrence frequency.
3. network text classification judges and cluster analysis: its objective is that the content of text Network Based adopts Text Classification to judge the generic of network text.Generic is the one in the public sentiment taxonomic hierarchies model set up in advance based on ontology, public sentiment taxonomic hierarchies model as shown in Figure 2, it is a hierarchical model, ground floor is large class, the second layer is group, each group is defined by concept attribute, as shown in Figure 3, has two concept attributes: classification semantic feature and prevention and control strategy.Classification semantic feature comprises:
Single word feature: the single word feature of the network text that classification semantic feature abstraction module extracts;
Word association feature: many word associations feature of the network text that classification semantic feature abstraction module extracts;
Geographic location feature: the geographic position name in the network text that classification semantic feature abstraction module extracts;
Character features: the person names in the network text that classification semantic feature abstraction module extracts;
Example: an example text of the type network public-opinion;
Classification judgment criterion.Judge whether the text accumulation that certain class public sentiment a collection of is relevant is really a public sentiment event.Such as, IF occurs that the Websites quantity of public sentiment text is greater than n THEN is a public sentiment event; It is a public sentiment event that the money order receipt to be signed and returned to the sender quantity of IF public sentiment text is greater than n THEN.
Prevention and control strategy comprises prevention principle and preventing control method, and prevention principle is the cardinal rule of carrying out defence for certain class public sentiment event and controlling; Preventing control method is the concrete prevention and control measure taked for certain class public sentiment.
Fig. 4 is the schematic diagram of an actual public sentiment taxonomic hierarchies.
Each classification has its category feature, for each classification produces the method for category feature as shown in Figure 5: first gather some network texts of each classification as training sample, the Chinese word cutting method of natural language processing technique is utilized to carry out participle and part-of-speech tagging process to all training samples, mark out the part of speech of each lexical item, give up to fall the vocabulary except noun, verb, adjective in text; Extract the single word feature of each text, word association feature, geographic location feature and character features by network text semantic feature extract function unit, then extract classification semantic feature by classification semantic feature extract function unit; Concrete grammar is: each feature utilizing computing machine to adopt statistic algorithm to calculate each text is in each classification and the Statistical Distribution of training sample complete or collected works, select to occur in classification sample files over half and be not vocabulary that in training sample complete or collected works, all samples are common as Based on Class Feature Word Quadric, and calculate the interior on average occurrence frequency of its classification as the frequency of Based on Class Feature Word Quadric.Generally speaking category feature is one group of vocabulary representing category feature, is furnished with its average occurrence frequency.
The method that network text classification judges is compared with each category feature lexical item by the feature lexical item of network text to mate, as shown in Figure 6, matching operation is carried out respectively in single word feature, word association feature, geographic location feature and character features four, and according to formulae discovery Similarity value below, text is attributed to the highest classification of Similarity value.
Wherein,
D represents document to be sorted;
C represents classification;
Coord (d, C) represents the quantity of the category feature lexical item comprising classification C in text d to be identified;
the word frequency of frequency representation feature lexical item t in category feature;
Weight (t): the weight of representation feature lexical item t;
Obtain in the category feature lexical item table that frequency and weight value can create from modeling process, category feature lexical item table is as shown in table 1.
Table 1 category feature lexical item table
| Classification | Feature Words | Word frequency | Weight |
| varchar | varchar | float | float |
NumofClasses: represent total several classification;
ClassFreq (t): representation feature item item t is the feature lexical item of several classification simultaneously.
As shown in Figure 7, network text is after preprocessing function cell processing, obtain text word segmentation result and remove stop words, its semantic feature is obtained again by semantic feature abstraction module, whether it is the one of known n kind network public-opinion to utilize the interpretation of network text classification arbitration functions unit, if then sorted out, otherwise, be given to network text cluster analysis functional unit to analyze, see wherein whether have much-talked-about topic, come each network text to collection and carry out classification judgement, the network text meeting class condition is composed with corresponding class label.If at the appointed time in section, belong to the quantity of the network text of a certain classification, occur that the Websites quantity of this classification network text exceedes the threshold value of specifying, then send alarm to system operators, and then provide emergence treatment scheme by emergence treatment scheme generation module.
In above-mentioned network text classification deterministic process, there will be the text that some do not belong to any class in existing public sentiment taxonomic hierarchies model, As time goes on, UNKNOWN TYPE text can constantly be accumulated, cluster analysis is carried out to the UNKNOWN TYPE text of accumulation, network text close for content is polymerized to bunch, if network amount of text exceeds certain threshold value in bunch, much-talked-about topic is then it can be used as to submit artificial interpretation to, if determine that it is new public sentiment classification, then public sentiment classification semantic feature is carried out to it and take out process, and the classification semantic feature of extraction is added in knowledge base, detailed process as shown in Figure 7, said process ensure that the extensibility of the knowledge base of native system, makes system can identify novel public sentiment on internet after supplementary knowledge.
4. emergence treatment scheme generates: be on the basis of public sentiment type identification, emergency disposal prediction scheme is provided for the public sentiment type identified, it is characterized in that, utilize internet public feelings event sight ontology knowledge library model and the network public-opinion prevention and control measure prediction scheme ontology knowledge library model of ontology technique construction stratification.The former carries out the description of quantitative and qualitative analysis to public sentiment event, as shown in Figure 8; The public sentiment that natural language mode word exists met an urgent need prevention and control rules and regulations, processing specification, counter-measure of the latter carries out digitizing, as shown in Figure 9.The object done like this changes the information of unformatted into computing machine intelligible formatted message.There is the support of above-mentioned two knowledge base models, just semantic matches technology can have been utilized automatically to realize the automatic identification of public sentiment event based on computing machine, the fast automatic reasoning of the corresponding precautionary measures, processing scheme, the real-time Aided Generation of process prediction scheme.Sight ontology knowledge base comprises the knowledge concepts such as public sentiment, time, website, participant, audient, potential hazard.
The information of the public sentiment event identified in internet information analysis and network text classification determining step can be extracted out and be stored in public sentiment event sight ontology knowledge base; Public sentiment classification information is provided by network text classification determining step, and what specifically adopt is Text Classification; Public sentiment content, time time of origin, time remaining time, web site name, Websites quantity, participant's user name are provided by internet information analytical procedure, employing be rule-based information extraction technique; Out of Memory such as the information such as public sentiment grade, participant IP address are then filled according to priori.
Public sentiment prevention and control measure prediction scheme ontology knowledge base comprises basis of compilation, the scope of application, resource, prevention and control measure four aspects, and its content is filled according to concrete laws and regulations content.
Together constitute network public-opinion prevention and control Knowledge Semantic Model Based based on internet public feelings event sight ontology knowledge base and network public-opinion prevention and control measure prediction scheme ontology knowledge base, based on this model, utilize semantic matches technology to generate emergency preplan, as shown in Figure 10.Emergency preplan instructs the scheme and method of disposing various public sentiment accident, and the actual conditions of each public sentiment event, situation and parameter are different, decision maker needs from prevention and control prediction scheme, to select suitable prevention and control Disposal Measures, method and implementation step as the case may be as emergency preplan, and allocates corresponding organizational structure and department's execution emergency preplan.For this reason, " the public sentiment classification ", " public sentiment content ", " public sentiment grade " of event sight is matched with prediction scheme body " being suitable for event type ", " being suitable for event content ", " being suitable for event class " respectively, as shown in Figure 11, thus find and the matched contingency plan of public sentiment event, as shown in table 2 and table 3.
The prediction scheme example that table 2 generates based on semantic matches
The explanation of table 3 prediction scheme example
Contingency plan is a guiding scheme, and need again according to the concrete condition of public sentiment, such as, the situations such as time, website, participant, audient, potential hazard generate concrete carrying into execution a plan.
5. emergency processing recruitment evaluation: emergency processing recruitment evaluation completes based on evaluation index system and evaluates calculation formula, evaluation index system contains the item needing assessment, and evaluates calculation formulae discovery goes out to quantize assessment result; As shown in Figure 12, the detailed description of each index is as shown in table 4 for evaluation index system.
Table 4 emergency processing recruitment evaluation index system
Public sentiment intensity index is intended to weigh public sentiment in scope and pro forma situation.1. public sentiment scope refers to the range of public sentiment, is weighed by website coverage, regional coverage degree, Websites quantity three indexs.Website coverage refers to the proportion that the website comprising public sentiment text accounts for sample site measure; Sample site measure is through to be chosen meticulously, can represent the set of websites of whole network state and level to a certain extent; Because the scale-level of each website is different, process to be weighted to it, occur that the sample site measure of public sentiment text is more, illustrate that the scope of public sentiment is wider, when after enforcement prevention and control measure, if the Websites quantity comprising public sentiment text occurs that the trend reduced illustrates that prevention and control measure has played effect.Regional coverage degree refers to the geographic distribution situation of the website comprising public sentiment text, occurs that the website distribution of public sentiment text is wider, illustrates that the coverage of public sentiment is wider.Websites quantity refers to the total quantity of the website comprising public sentiment text, and quantity is more, illustrates that the coverage of public sentiment is wider.2. public sentiment form refers to media channel kind, the length of network text used, the medium kind of network text that public sentiment is propagated.Media channel kind can be BBS, microblogging, blog, friend-making platform, Email etc., and channel used is more, then transmission capacity is stronger.The length of network text used is longer, then transmission capacity is stronger.Medium kind can be text, audio frequency, video, and public sentiment impact is stronger more at most for medium used thereof kind.
Audient's attention rate index is intended to reflection network public-opinion to the influence power of audient, is weighed by indexs such as audient's situation, audient's response, audient's attitudes.1. audient's situation refers to the audience size and audient's scope that affect by public sentiment, and audience size is measured by network text viewer IP quantity, and audient's scope is measured by the distributional region range of network text viewer IP.2. audient's response refers to the degree of concern of viewer to network text, is weighed by amount of reading, transfer amount, money order receipt to be signed and returned to the sender amount, liveness.Amount of reading is measured by the touching quantity of network text, transfer amount is measured by the occurrence number of network text different web sites within the scope of full internet, money order receipt to be signed and returned to the sender amount is replied quantity by network text and is measured, liveness refers to the degree of recognition of viewer to the viewpoint expressed by network text by measuring 3. audient's attitude to the reply quantity of network text in the unit interval, is weighed by front attitude money order receipt to be signed and returned to the sender quantity, middle sexual attitude money order receipt to be signed and returned to the sender quantity, negative attitude money order receipt to be signed and returned to the sender quantity.
The weight of the indexs at different levels of this index system is calculated by analytical hierarchy process, and each index all can draw by quantum chemical method, and the Quantitative Calculation Method of index is divided into three kinds: index calculating, frequency/density calculation and weight coefficient are determined.
(1) index calculates
Quantitative target and qualitative index is had in index system.Quantitative target comprises the indexs such as amount of reading, transfer amount, money order receipt to be signed and returned to the sender amount; Qualitative index comprises audiovisual degree.For having comparability, qualitative index and quantitative target being pressed normalized, adopts index calculation method here, concrete employing Sigmoid function
calculate, wherein x represents amount of reading, transfer amount, money order receipt to be signed and returned to the sender amount etc.For audient's response, if for network text i, the touching quantity of network text is x
1i, the occurrence number of network text different web sites within the scope of full internet is x
2i, it is x that network text replys quantity
3i, the unit interval is interior is x to the reply quantity of network text
4i.If the weight of amount of reading, transfer amount, money order receipt to be signed and returned to the sender amount, liveness is g1, g2, g3, g4, then network text to the influence power P1 that audient responds is:
P1＝f(x
1i)×g
1+f(x
2i)×g
2+f(x
3i)×g
3+f(x
4i)×g
4
(2) frequency computation part
Liveness weighs the reply frequency of network text according to netizen, with sky, week, the moon for timing statistics unit.
(3) weight coefficient is determined
Analytical hierarchy process is utilized to determine the weight coefficient of each attribute factor according to expertise.Its principal character be the PROBLEM DECOMPOSITION of complexity for several compositing factors, these factors are divided into hierarchical structure by subordinate relation; Only need compare between two each factor when expert appraises through comparison, determine the relative importance of factors in same level, then the judgement of comprehensive expert determines the order that each factor is relatively important.The weighting coefficient deciding each factor in this way makes weighting coefficient by rule of thumb simultaneously than in several factors more more scientific, judges more accurately because easily draw when people only compare between two.But when using these methods, in order in order to be effective, the factor that each level comprises is generally more than 10.Undertaken by 9 points of systems when contrasting between two, quite, 3 is slightly good, and 5 is good significantly, and 7 is very good, and 9 is fabulous in 1 representative.As then represented with 2,4,6 or 8 points between said two devices.Forming rating matrix according to contrasting marking result between two, each factor can be calculated relative to the importance of last layer target or evaluation weight by asking the Maximum characteristic root of matrix and proper vector.If require to calculate each parameter to the sequence of importance of last layer target again or influence degree size, the weight of each parameter of bottom can be multiplied by one by one the weight of the last layer factor relevant with it, then be added, each like this parameter has just been calculated the order of quality of last layer again or weighting coefficient.
The computing formula of quantitative evaluation result is,
Wherein, A
irepresent first class index, the score value of public sentiment intensity and audient's attention rate, ω
irepresent respective weight.
Each first class index is then determined by the two-level index of its subordinate, and computing formula is
wherein,
be the jth item of i-th first class index, its weight is ω
j.Similarly, each two-level index is determined by three grades of indexs of its subordinate.
Claims (6)
1. an intelligent public sentiment accident emergent treatment system, is characterized in that: this system comprises internet information acquisition and parsing module, internet information analysis module, network text classification judge and Cluster Analysis module, emergence treatment scheme generation module and emergency processing recruitment evaluation module; Described internet information acquisition and parsing module, for from Information Monitoring on internet, extract the metadata information of natural language word and webpage in webpage, and are saved in database; The natural language word that described internet information analysis module is used for gathering in next information carries out feature extraction, forms text feature; Described network text classification judges to be used for judging the classification of network text with Cluster Analysis module, carries out cluster analysis to cumulative network text; Described emergence treatment scheme generation module is used for automatically generating according to the concrete condition of public sentiment event processing prediction scheme accordingly, and decision-maker can based on process prediction scheme formulation and implementation scheme; Described emergency processing recruitment evaluation module is used for assessing the implementation effect carried into execution a plan.
2. an intelligent public sentiment accident emergency processing method, is characterized in that comprising the following steps:
1. internet information acquisition and parsing: by the computing machine of connecting Internet from collection network text internet; Then, computing machine adopts rule-based information extraction technique automatically to resolve network text, from wherein extracting two category informations: the metadata information of natural language Word message and webpage;
2. internet information analysis: first utilize the Chinese word cutting method of natural language processing technique to carry out participle respectively to the title of network text and body matter, and the part of speech of lexical item each in word segmentation result is marked, give up to fall the lexical item except noun, verb, adjective afterwards, then utilize text multiple-accuracy representing method to extract the single lexical item characteristic sum lexical item linked character of network text, then identify geographic location feature in network text and character features according to the part-of-speech tagging situation in word segmentation result;
3. the lexical item in the network text after step 2. being processed is compared with the lexical item feature of the public sentiment classification set in Computer Database and is mated, and according to matching result, network text is carried out classification process according to the public sentiment classification set in Computer Database; The network text that can not sort out carries out cluster analysis, network text close for content is polymerized to bunch, if network amount of text exceeds setting threshold value in bunch, then to bunch in the network text lexical item feature of carrying out public sentiment classification take out process, and the lexical item feature of the public sentiment classification of extraction is added in Computer Database; 4. step is proceeded to for the network text completing classification; Wherein, matching content comprises single lexical item feature, lexical item linked character, geographic location feature and character features;
If 4. at the appointed time in section, belong to the quantity of the network text of a certain classification or occur that the Websites quantity of this classification network text exceedes the threshold value of specifying, then initiate emergency plan;
Complete the emergency processing of intelligent public sentiment accident.
3. the public sentiment accident emergency processing method of a kind of intelligence according to claim 2, it is characterized in that: after step 4., also comprise emergency processing recruitment evaluation step: first according to evaluation index acquisition index data, then achievement data input assessment formula is drawn quantitative evaluation result.
4. the public sentiment accident emergency processing method of a kind of intelligence according to claim 2, it is characterized in that: step 3. according to matching result by network text according to Computer Database in the public sentiment classification that sets carry out sorting out process and be specially: the method that network text classification judges the lexical item of network text is compared with the lexical item feature of each public sentiment classification to mate, respectively in single word feature, word association feature, matching operation is carried out in geographic location feature and character features four aspects, the Similarity value of network text and each public sentiment classification is obtained according to match condition, text is attributed to the public sentiment classification that Similarity value is the highest.
5. the public sentiment accident emergency processing method of a kind of intelligence according to claim 2, is characterized in that: step 3. in bunch in the network text lexical item feature of carrying out public sentiment classification take out process, be specially: suppose that the network text that bunch T comprises has T={t
1, t
2... t
n, utilize text multiple-accuracy representing method to extract each text t
isingle lexical item characteristic sum lexical item linked character, statistical method is adopted to calculate the Statistical Distribution of all single lexical item characteristic sum lexical item linked character of all texts in T again, select the vocabulary that occurred in network text over half in T as public sentiment classification lexical item feature, and calculate the frequency of its average occurrence frequency in T as public sentiment category feature lexical item; Wherein, 1≤i≤n.
6. the public sentiment accident emergency processing method of a kind of intelligence according to claim 2, it is characterized in that: step 4. in the generation method of emergency preplan be: based on internet public feelings event sight ontology knowledge library model and network public-opinion prevention and control measure prediction scheme ontology knowledge base, utilize semantic matches technology according to the concrete condition of public sentiment event sight, from prevention and control measure prediction scheme storehouse, Auto-matching goes out optimal plan for emergency handling.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510243751.0A CN104820629B (en) | 2015-05-14 | 2015-05-14 | A kind of intelligent public sentiment accident emergent treatment system and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201510243751.0A CN104820629B (en) | 2015-05-14 | 2015-05-14 | A kind of intelligent public sentiment accident emergent treatment system and method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104820629A true CN104820629A (en) | 2015-08-05 |
| CN104820629B CN104820629B (en) | 2018-01-30 |
Family
ID=53730930
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201510243751.0A Active CN104820629B (en) | 2015-05-14 | 2015-05-14 | A kind of intelligent public sentiment accident emergent treatment system and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104820629B (en) |
Cited By (29)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105956740A (en) * | 2016-04-19 | 2016-09-21 | 北京深度时代科技有限公司 | Semantic risk calculating method based on text logical characteristic |
| CN106202561A (en) * | 2016-07-29 | 2016-12-07 | 北京联创众升科技有限公司 | Digitized contingency management case library construction methods based on the big data of text and device |
| CN106294619A (en) * | 2016-08-01 | 2017-01-04 | 上海交通大学 | Public sentiment intelligent supervision method |
| CN107193796A (en) * | 2016-03-14 | 2017-09-22 | 北大方正集团有限公司 | A kind of public sentiment event detecting method and device |
| CN107239452A (en) * | 2016-03-28 | 2017-10-10 | 腾讯科技（深圳）有限公司 | A kind of method and device of Developing Tactics |
| CN107274324A (en) * | 2017-06-06 | 2017-10-20 | 张黎明 | A kind of method that accident risk assessment is carried out based on cloud service |
| CN107491438A (en) * | 2017-08-25 | 2017-12-19 | 前海梧桐（深圳）数据有限公司 | Business decision elements recognition method and its system based on natural language |
| CN107590196A (en) * | 2017-08-15 | 2018-01-16 | 中国农业大学 | Earthquake emergency information screening and evaluating system and system in a kind of social networks |
| CN107622354A (en) * | 2017-09-29 | 2018-01-23 | 中国科学技术大学 | A kind of accident contingency capabilities assessment method based on section Two-tuple Linguistic Information Processing |
| CN107741929A (en) * | 2017-10-18 | 2018-02-27 | 网智天元科技集团股份有限公司 | The analysis of public opinion method and device |
| CN108108902A (en) * | 2017-12-26 | 2018-06-01 | 阿里巴巴集团控股有限公司 | A kind of risk case alarm method and device |
| CN108255832A (en) * | 2016-12-28 | 2018-07-06 | 航天信息股份有限公司 | public sentiment processing system and method |
| CN109615266A (en) * | 2018-12-26 | 2019-04-12 | 贵州电网有限责任公司 | The text analyzing decision-making technique of power grid exception information based on data mining |
| CN110046220A (en) * | 2018-12-13 | 2019-07-23 | 阿里巴巴集团控股有限公司 | Public feelings information processing method, device, equipment and computer readable storage medium |
| CN110096406A (en) * | 2018-01-31 | 2019-08-06 | 阿里巴巴集团控股有限公司 | A kind of event of failure discovery method and server |
| CN110609969A (en) * | 2019-08-08 | 2019-12-24 | 阿里巴巴集团控股有限公司 | Information processing method and device |
| CN110852090A (en) * | 2019-11-07 | 2020-02-28 | 中科天玑数据科技股份有限公司 | Public opinion crawling mechanism characteristic vocabulary extension system and method |
| CN110868383A (en) * | 2018-12-24 | 2020-03-06 | 北京安天网络安全技术有限公司 | Website risk assessment method and device, electronic equipment and storage medium |
| CN111223026A (en) * | 2020-01-03 | 2020-06-02 | 武汉理工大学 | Intelligent management method for garbage crisis conversion |
| CN111428146A (en) * | 2020-03-24 | 2020-07-17 | 上海智臻智能网络科技股份有限公司 | Network information processing method and system, equipment and storage medium |
| CN111898385A (en) * | 2020-07-17 | 2020-11-06 | 中国农业大学 | Earthquake disaster assessment method and system |
| CN112069381A (en) * | 2020-09-27 | 2020-12-11 | 中国科学院深圳先进技术研究院 | Monitoring management method and system based on natural language processing technology |
| CN112905745A (en) * | 2021-03-05 | 2021-06-04 | 广州虎牙科技有限公司 | Information processing method, information processing device, electronic equipment and computer readable storage medium |
| WO2021136453A1 (en) * | 2019-12-31 | 2021-07-08 | 清华大学 | Method and apparatus for obtaining emergency plan auxiliary information, and device |
| CN113433994A (en) * | 2021-07-20 | 2021-09-24 | 青岛一云数联科技有限公司 | System and method for sensing and monitoring energy event |
| CN113468320A (en) * | 2021-07-22 | 2021-10-01 | 中国地震台网中心 | Method and system for quickly visualizing earthquake emergency information |
| CN113626722A (en) * | 2020-05-08 | 2021-11-09 | 国家广播电视总局广播电视科学研究院 | Public opinion guiding method, device, equipment and computer readable storage medium |
| CN114417021A (en) * | 2022-01-24 | 2022-04-29 | 中国电子科技集团公司第五十四研究所 | Semantic information accurate distribution method based on time, space and semantic multi-constraint fusion |
| CN114444514A (en) * | 2022-02-08 | 2022-05-06 | 北京百度网讯科技有限公司 | Semantic matching model training method, semantic matching method and related device |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101436194A (en) * | 2008-11-04 | 2009-05-20 | 中国电子科技集团公司第五十四研究所 | Text multiple-accuracy representing method based on data excavating technology |
| CN101819573A (en) * | 2009-09-15 | 2010-09-01 | 电子科技大学 | Self-adaptive network public opinion identification method |
| CN102509164A (en) * | 2011-11-24 | 2012-06-20 | 广州市地下铁道总公司 | Automatic generation method for digital emergency plan |
| CN103150335A (en) * | 2013-01-25 | 2013-06-12 | 河南理工大学 | Co-clustering-based coal mine public sentiment monitoring system |
| CN103544255A (en) * | 2013-10-15 | 2014-01-29 | 常州大学 | Text semantic relativity based network public opinion information analysis method |
| CN104573016A (en) * | 2015-01-12 | 2015-04-29 | 武汉泰迪智慧科技有限公司 | System and method for analyzing vertical public opinions based on industry |
-
2015
- 2015-05-14 CN CN201510243751.0A patent/CN104820629B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101436194A (en) * | 2008-11-04 | 2009-05-20 | 中国电子科技集团公司第五十四研究所 | Text multiple-accuracy representing method based on data excavating technology |
| CN101819573A (en) * | 2009-09-15 | 2010-09-01 | 电子科技大学 | Self-adaptive network public opinion identification method |
| CN102509164A (en) * | 2011-11-24 | 2012-06-20 | 广州市地下铁道总公司 | Automatic generation method for digital emergency plan |
| CN103150335A (en) * | 2013-01-25 | 2013-06-12 | 河南理工大学 | Co-clustering-based coal mine public sentiment monitoring system |
| CN103544255A (en) * | 2013-10-15 | 2014-01-29 | 常州大学 | Text semantic relativity based network public opinion information analysis method |
| CN104573016A (en) * | 2015-01-12 | 2015-04-29 | 武汉泰迪智慧科技有限公司 | System and method for analyzing vertical public opinions based on industry |
Non-Patent Citations (1)
| Title |
|---|
| 马力: "基于Web2.0的政府网络舆情信息管理模型研究", 《中国优秀硕士学位论文全文数据库 信息科技辑》 * |
Cited By (40)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107193796A (en) * | 2016-03-14 | 2017-09-22 | 北大方正集团有限公司 | A kind of public sentiment event detecting method and device |
| CN107193796B (en) * | 2016-03-14 | 2021-12-24 | 北大方正集团有限公司 | Public opinion event detection method and device |
| CN107239452A (en) * | 2016-03-28 | 2017-10-10 | 腾讯科技（深圳）有限公司 | A kind of method and device of Developing Tactics |
| CN105956740A (en) * | 2016-04-19 | 2016-09-21 | 北京深度时代科技有限公司 | Semantic risk calculating method based on text logical characteristic |
| CN105956740B (en) * | 2016-04-19 | 2019-12-31 | 北京深度时代科技有限公司 | Semantic risk calculation method based on text logical features |
| CN106202561A (en) * | 2016-07-29 | 2016-12-07 | 北京联创众升科技有限公司 | Digitized contingency management case library construction methods based on the big data of text and device |
| CN106202561B (en) * | 2016-07-29 | 2019-10-01 | 北京联创众升科技有限公司 | Digitlization contingency management case base construction method and device based on text big data |
| CN106294619A (en) * | 2016-08-01 | 2017-01-04 | 上海交通大学 | Public sentiment intelligent supervision method |
| CN108255832A (en) * | 2016-12-28 | 2018-07-06 | 航天信息股份有限公司 | public sentiment processing system and method |
| CN107274324A (en) * | 2017-06-06 | 2017-10-20 | 张黎明 | A kind of method that accident risk assessment is carried out based on cloud service |
| CN107590196A (en) * | 2017-08-15 | 2018-01-16 | 中国农业大学 | Earthquake emergency information screening and evaluating system and system in a kind of social networks |
| CN107491438A (en) * | 2017-08-25 | 2017-12-19 | 前海梧桐（深圳）数据有限公司 | Business decision elements recognition method and its system based on natural language |
| CN107622354A (en) * | 2017-09-29 | 2018-01-23 | 中国科学技术大学 | A kind of accident contingency capabilities assessment method based on section Two-tuple Linguistic Information Processing |
| CN107622354B (en) * | 2017-09-29 | 2020-06-26 | 中国科学技术大学 | Emergency capacity evaluation method for emergency events based on interval binary semantics |
| CN107741929A (en) * | 2017-10-18 | 2018-02-27 | 网智天元科技集团股份有限公司 | The analysis of public opinion method and device |
| CN108108902A (en) * | 2017-12-26 | 2018-06-01 | 阿里巴巴集团控股有限公司 | A kind of risk case alarm method and device |
| CN108108902B (en) * | 2017-12-26 | 2021-06-29 | 创新先进技术有限公司 | Risk event warning method and device |
| CN110096406A (en) * | 2018-01-31 | 2019-08-06 | 阿里巴巴集团控股有限公司 | A kind of event of failure discovery method and server |
| CN110046220A (en) * | 2018-12-13 | 2019-07-23 | 阿里巴巴集团控股有限公司 | Public feelings information processing method, device, equipment and computer readable storage medium |
| CN110868383A (en) * | 2018-12-24 | 2020-03-06 | 北京安天网络安全技术有限公司 | Website risk assessment method and device, electronic equipment and storage medium |
| CN109615266A (en) * | 2018-12-26 | 2019-04-12 | 贵州电网有限责任公司 | The text analyzing decision-making technique of power grid exception information based on data mining |
| CN109615266B (en) * | 2018-12-26 | 2022-11-04 | 贵州电网有限责任公司 | Text analysis decision method for power grid abnormal information based on data mining |
| CN110609969A (en) * | 2019-08-08 | 2019-12-24 | 阿里巴巴集团控股有限公司 | Information processing method and device |
| CN110852090A (en) * | 2019-11-07 | 2020-02-28 | 中科天玑数据科技股份有限公司 | Public opinion crawling mechanism characteristic vocabulary extension system and method |
| CN110852090B (en) * | 2019-11-07 | 2024-03-19 | 中科天玑数据科技股份有限公司 | Mechanism characteristic vocabulary expansion system and method for public opinion crawling |
| WO2021136453A1 (en) * | 2019-12-31 | 2021-07-08 | 清华大学 | Method and apparatus for obtaining emergency plan auxiliary information, and device |
| CN111223026A (en) * | 2020-01-03 | 2020-06-02 | 武汉理工大学 | Intelligent management method for garbage crisis conversion |
| CN111223026B (en) * | 2020-01-03 | 2024-03-01 | 武汉理工大学 | Intelligent management method for garbage crisis transformation |
| CN111428146A (en) * | 2020-03-24 | 2020-07-17 | 上海智臻智能网络科技股份有限公司 | Network information processing method and system, equipment and storage medium |
| CN113626722A (en) * | 2020-05-08 | 2021-11-09 | 国家广播电视总局广播电视科学研究院 | Public opinion guiding method, device, equipment and computer readable storage medium |
| CN111898385B (en) * | 2020-07-17 | 2023-08-04 | 中国农业大学 | Earthquake disaster assessment method and system |
| CN111898385A (en) * | 2020-07-17 | 2020-11-06 | 中国农业大学 | Earthquake disaster assessment method and system |
| CN112069381A (en) * | 2020-09-27 | 2020-12-11 | 中国科学院深圳先进技术研究院 | Monitoring management method and system based on natural language processing technology |
| CN112905745A (en) * | 2021-03-05 | 2021-06-04 | 广州虎牙科技有限公司 | Information processing method, information processing device, electronic equipment and computer readable storage medium |
| CN113433994A (en) * | 2021-07-20 | 2021-09-24 | 青岛一云数联科技有限公司 | System and method for sensing and monitoring energy event |
| CN113468320A (en) * | 2021-07-22 | 2021-10-01 | 中国地震台网中心 | Method and system for quickly visualizing earthquake emergency information |
| CN114417021A (en) * | 2022-01-24 | 2022-04-29 | 中国电子科技集团公司第五十四研究所 | Semantic information accurate distribution method based on time, space and semantic multi-constraint fusion |
| CN114417021B (en) * | 2022-01-24 | 2023-08-25 | 中国电子科技集团公司第五十四研究所 | Semantic information accurate distribution method based on time, space and sense multi-constraint fusion |
| CN114444514A (en) * | 2022-02-08 | 2022-05-06 | 北京百度网讯科技有限公司 | Semantic matching model training method, semantic matching method and related device |
| CN114444514B (en) * | 2022-02-08 | 2023-01-24 | 北京百度网讯科技有限公司 | Semantic matching model training method, semantic matching method and related device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104820629B (en) | 2018-01-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104820629B (en) | A kind of intelligent public sentiment accident emergent treatment system and method | |
| CN107515873B (en) | Junk information identification method and equipment | |
| Jiang et al. | Public-opinion sentiment analysis for large hydro projects | |
| Rubin et al. | Veracity roadmap: Is big data objective, truthful and credible? | |
| CN106991085B (en) | Entity abbreviation generation method and device | |
| Kim et al. | Crystal: Analyzing predictive opinions on the web | |
| CN102591854B (en) | For advertisement filtering system and the filter method thereof of text feature | |
| CN106611375A (en) | Text analysis-based credit risk assessment method and apparatus | |
| CN110674840B (en) | Multi-party evidence association model construction method and evidence chain extraction method and device | |
| CN109145216A (en) | Network public-opinion monitoring method, device and storage medium | |
| CN103399891A (en) | Method, device and system for automatic recommendation of network content | |
| Sharma et al. | Nlp and machine learning techniques for detecting insulting comments on social networking platforms | |
| CN107291723A (en) | The method and apparatus of web page text classification, the method and apparatus of web page text identification | |
| CN111967761A (en) | Monitoring and early warning method and device based on knowledge graph and electronic equipment | |
| CN112419029B (en) | Similar financial institution risk monitoring method, risk simulation system and storage medium | |
| CN107688870A (en) | A kind of the classification factor visual analysis method and device of the deep neural network based on text flow input | |
| CN113342972B (en) | Public opinion recognition model training method and system and public opinion risk monitoring method and system | |
| Samonte et al. | Sentence-level sarcasm detection in English and Filipino tweets | |
| Samonte | Polarity analysis of editorial articles towards fake news detection | |
| CN104794209A (en) | Chinese microblog sentiment classification method and system based on Markov logic network | |
| CN114764463A (en) | Internet public opinion event automatic early warning system based on event propagation characteristics | |
| CN109658148B (en) | Marketing activity complaint risk prediction method based on natural language processing technology | |
| Ruichen | The Basic Principles of Marxism with the Internet as a Carrier | |
| CN111767404B (en) | Event mining method and device | |
| Drury | A Text Mining System for Evaluating the Stock Market's Response To News |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| EXSB | Decision made by sipo to initiate substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |