CN103198078B - A kind of internet news event report trend analysis and system - Google Patents
A kind of internet news event report trend analysis and system Download PDFInfo
- Publication number
- CN103198078B CN103198078B CN201210004878.3A CN201210004878A CN103198078B CN 103198078 B CN103198078 B CN 103198078B CN 201210004878 A CN201210004878 A CN 201210004878A CN 103198078 B CN103198078 B CN 103198078B
- Authority
- CN
- China
- Prior art keywords
- theme
- information
- news
- subject
- internet
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
Landscapes
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention discloses a kind of internet news event report trend analysis and system, relate to internet information analysis field.The analysis method of existing media event development trend can not be understood the subject information of media event or can not screen its Data Source.The method of the invention and system are with user's media event of interest as starting point, the news information that user in the Internet may be concerned about by the characteristic information according to the media event arranged is acquired and carries out Preliminary screening, then the theme drawing media event is analyzed, and measure out conversion theme according to the theme in different cycles and related information, finally the distribution trend that each theme and relevant report quantity thereof develop in time is represented.By the method and system, user can grasp development trend and the propagation law of internet news event more intuitively and easily.
Description
Technical field
The present invention relates to internet information analysis technical field, be specifically related to a kind of internet news event report trend analysis and system.
Background technology
The generation of the Internet brings the revolutionary leap of mankind's circulation way, is becoming the basic communication instrument of information-intensive society.Especially when important event occurs, increasing people first passes through the Internet and obtains information, gives opinion, and the Internet has become the distribution centre of various information and very important important public opinion position.
As a example by Beijing, according to the preliminary statistics, before 2011, the nearlyest 260,000 of the website quantity of the Internet, Beijing area；Netizen's number reaches 4,020,000 people；Current Beijing area relates to the website of Internet news communications and about publishes news 30,000, more new post literary composition 300,000, transmission SMS news 3,000 every day；The page browsing amount of all websites every days reaches 1,000,000,000 page numbers, visitor more than 100,000,000 person-times.
In sum, internet information is explosive growth, and internet information is increasingly paid close attention to by government, enterprise, the user of individual's every aspect, also is intended to find required information quickly and efficiently in the Internet simultaneously.Topic detection technology can analyze theme in real time from magnanimity information, and such user just by browsing theme and analyzing, thus can understand the content in magnanimity information.
Analysis method to known news event development trend at present, a kind of development trend being to realize analyzing media event by global search technology, this kind of method cannot understand the subject information of event.Another kind is to be realized by topic detection technology, the Data Source of this kind of method is defined in the information such as news, forum, blog at present, Data Source is not carried out further screening operation, and in analysis, also do not embody the key topic causing theme to divide, and relevant representative hot news information.
Summary of the invention
For defect present in prior art, the technical problem to be solved is to provide a kind of internet news event report trend analysis and system, the method and system can show, by theme and relevant report number change, the distribution trend that media event develops in time, and by the hot news theme in thematic relation, reprinting amount and comment number analysis distribution trend.
In order to solve above-mentioned technical problem, the technical solution used in the present invention is as follows:
A kind of internet news event report trend analysis, comprises the following steps:
(1) according to the characteristic information of the media event configured, news information in Real-time Collection the Internet；
(2) news information described in pretreatment, filters out issuing time news information in the measurement period of the media event arranged；
(3) text message of the news information that Analysis and Screening goes out, obtains theme and the subject information of news information；Described theme includes the theme that there is subject information and there is not the source theme of subject information；
(4) analyze the theme in current statistic cycle and previous measurement period, draw the related information number of the same subject between current statistic cycle and previous measurement period and each same subject；
(5) according to described related information number, conversion theme is drawn；Described conversion theme refers to the related information number theme more than or equal to the related information number of previous this same subject of measurement period of this same subject of current statistic cycle；
(6) obtain the source theme set in the time period and convert the related information number of theme, and represent source theme and convert the related information number of theme distribution trend in time.
A kind of internet news event report trend analysis system, including the information collecting device of news information in the characteristic information Real-time Collection the Internet according to media event；
For filtering out the data prediction device of issuing time news information in the measurement period of the media event arranged；
For the text message of the news information that Analysis and Screening goes out, obtain the theme of news information and the topic detection device of subject information；
For analyzing and draw the thematic relation analytical equipment of related information number of the same subject between current statistic cycle and previous measurement period and each same subject；
For according to the related information number of same subject between current statistic cycle and previous measurement period, drawing the conversion subject analysis device converting theme；
For obtaining and represent the distribution trend demonstration device of the related information number setting time period endogenous theme and convert theme distribution trend in time.
The method of the invention and system, based on the information gathering of internet information, news information cluster, information retrieval system, to the media event in known the Internet, calculate and analyze the related news theme of news thing, related news report reprinting amount and related commentary number thereof, show, by theme and relevant report number change, the distribution trend that media event develops in time；And by the theme of hot news in thematic relation, reprinting amount and comment number analysis distribution trend, and represent intuitively with graph mode, make user be best understood from, grasp the propagation law of internet news event.
Accompanying drawing explanation
Fig. 1 is the structured flowchart of internet news event report trend analysis system in detailed description of the invention；
Fig. 2 is the flow chart of mutual internet news event report trend analysis in detailed description of the invention.
Detailed description of the invention
Below in conjunction with Figure of description, the present invention is described in further detail with detailed description of the invention.
As it is shown in figure 1, internet news event report trend analysis system includes information collecting device 11, data prediction device 12, topic detection device 13, thematic relation analytical equipment 14, converts subject analysis device 15 and distribution trend demonstration device 16 in present embodiment.Wherein, data prediction device 12 includes that data Preliminary screening module 121 and data screen module 122 again.
Information collecting device 11 is for according to the news information in the characteristic information Real-time Collection the Internet of media event.Described characteristic information includes the logical relation between the key word of news information, keyword weight and key word；Described news information includes text message and review information.
Data prediction device 12 is for filtering out issuing time news information in the measurement period of set media event.The news information that data Preliminary screening module 121 is gathered in filtering out the measurement period that issuing time is media event.Data screen module 122 again for when the news information that data Preliminary screening module 121 is screened does not exists subject information, screening news information further according to characteristic information.
The text message of the news information that topic detection device 13 goes out for Analysis and Screening, obtains theme and the subject information of news information.
Thematic relation analytical equipment 14 is for analyzing and draw the related information number of the same subject between current statistic cycle and previous measurement period and each same subject.
Convert subject analysis device 15 to be used for according to the related information number of same subject between current statistic cycle and previous measurement period, draw conversion theme.
Distribution trend demonstration device 16 is for obtaining and representing the related information number setting time period endogenous theme and convert theme distribution trend in time.Representing source theme by scatterplot in present embodiment and convert the related information number distribution trend in time of theme, X-axis represents the time of theme, and Y-axis represents the related information number of theme.Distribution trend demonstration device represents the related information number of source theme distribution trend in time, represents including to source theme, the follow-up same subject of source theme, source theme relating subject and source theme subsequent association theme；Distribution trend demonstration device, to converting the representing of related information number distribution trend in time of theme, including to converting theme, converts the follow-up same subject of theme, converts representing of theme subsequent association theme.
As in figure 2 it is shown, the method for systematic analysis internet news event report trend comprises the following steps described in employing Fig. 1:
(1) information collecting device 11 is according to the characteristic information of media event, gathers the news information in the Internet.
The characteristic information of the media event according to configuration, news information in Real-time Collection the Internet.The characteristic information of media event includes key word, logical relation between the weight of key word and key word.In present embodiment, key word can arrange corresponding key word according to the related news time of understanding required for oneself for the kernel keyword of this media event, user.Keyword weight represents each key word status in overall crucial phrase.Logical relation between key word, is completed by logical relation expression formula, including or, non-and between nest relation.
The collection of news information includes that news category text message and news analysis information, collection include news information and downloads, analyzes, extracts key metadata item and preservation information.Wherein, the metadata of news information includes title, text, web site name, channel designation, issuing time；The metadata of news analysis information includes commenting on text, comment issuing time and comment author.The principle of selecting gathering information source is flow portal website big, influential.In present embodiment, the information source of collection includes the news channel of domestic tens of the portal websites such as Sina website, sohu.com, Netease's net, www.qq.com, the www.xinhuanet.com, People's Net.The wherein review information to news information, uses real-time incremental update mechanism, it is ensured that the review information of up-to-date issue can instantaneous acquiring.
Concrete acquisition method can use Chinese patent application document, Application No. 200910217505.2, entitled " a kind of method and device based on the Internet collecting webpage data of direction site ".
(2) data prediction device 12 pretreatment news information, filters out qualified news information.
The news information that pretreatment is gathered, filters out issuing time news information in the measurement period of set media event.The measurement period of media event set in present embodiment be the same day 16 when next day 9 because between during the morning 8～9, afternoon 15～16 time between be domestic main news portal website time period of updating news information.The 16 timing statistics cycles when next day 9 were set, can analyze to the greatest extent the news information of issue recently.Wherein, the method for news information screening is as follows:
1. data Preliminary screening module 121 filters out issuing time is the collection information in media event measurement period；
2. data are again screened module 122 and are checked during step is 1. whether the news information filtered out exists subject information, if there is then having screened, if there is no screening the most further, filter qualified news information by the logical relation between user-defined key word, key word.When screening qualified collection information further, only include body information, do not include news analysis information.
(3) topic detection device 13 analyzes the text of news information, obtains theme and subject information.
The text message of the news information that topic detection device 13 Analysis and Screening goes out, detection theme, form subject information and preserve, if there is not subject information in news information, then all themes are sorted from high to low by degree of association, retaining front 10 themes of ranking, labelling theme is the source theme of event, and preserves theme.
Degree of association between theme can use currently a popular and general calculating degree of association method, the most identical with the computational methods of Topic Similarity in step (4).
(4) theme in current statistic cycle and previous measurement period analyzed by thematic relation analytical equipment 14, draws same subject and related information number.
Analyze the theme in current statistic cycle and previous measurement period, draw the related information number of the same subject between current statistic cycle and previous measurement period and each same subject.Related information number refers within the current statistic cycle by the relevant information number in calculated same subject.The measurement period of the media event selected in present embodiment is two cycles, it is respectively (current statistic cycle) and the day before yesterday 16 when next day 9 (previous measurement period) when next day 9 on the same day 16, by calculating the similarity between current statistic cycle and the theme of previous measurement period, analyze the relating subject Information Number drawing same subject and each same period, it is marked and preserves, eliminate other subject information in the current statistic cycle.
Similarity between theme can use current general-purpose computations similarity based method Jaccard formula:
Wherein x and y represents the center vector of each theme respectively.
One of them Topic Similarity method: x of calculating current statistic theme and previous measurement period can represent the center vector of a theme in previous measurement period, y represents the center vector of a theme in current period, if the Similarity value obtained is higher than similarity threshold values set in advance, then it is assumed that two themes are correlated with.According to the method, the theme in current statistic cycle is calculated similarity with previous measurement period successively, obtain the related subject of current period theme and a theme in previous measurement period with this.
(5) conversion subject analysis device 15 is according to the related information number of same subject between current statistic cycle and previous measurement period, draws conversion theme.
Convert theme and refer to the related information number theme more than or equal to the related information number of previous this same subject of measurement period of this same subject of current statistic cycle.After drawing conversion theme, being ranked up the relevant information converted in theme from high to low by similarity, the highest information of labelling similarity is that it represents article.
Relevant information similarity in theme, is in the cluster process of the information of carrying out, calculates.Circular can use the most general text vector spatial model and similarity calculating method, i.e. represents text model by vector space, calculates the similarity between two texts eventually through the cosine law.
(6) distribution trend demonstration device 16 obtains the theme and related information number set in the time period and represents.
Obtain the source theme set in the time period and convert the related information number of theme, and represent source theme and convert the related information number of theme distribution trend in time.In present embodiment, representing the distribution trend of media event by the way of scatterplot, the X-axis of scatterplot is for representing the time of theme, and Y-axis is for representing the relevant information number of theme, and in scatterplot, available different colours or shape distinguish all kinds of themes.Theme can be divided mainly into two big class themes, and one type is source theme and related subject thereof, including source theme, the follow-up same subject of source theme, source theme relating subject and subsequent association theme.Wherein theme follow-up same subject in source refers to that the theme that n-th measurement period analysis draws is all the same subject of its previous measurement period, wherein n > 1；Source theme subsequent association theme refers to relating subject and the relating subject of the follow-up same subject of source theme of source theme.Source theme, the follow-up same subject of source theme, source theme subsequent association theme same color are expressed.
Another kind of for converting theme, converting the follow-up same subject of theme and convert the subsequent association theme same color expression of theme.Wherein convert the follow-up same subject of theme to refer to assume that n-th measurement period analysis draws and turn, change theme, thus convert the same subject that the theme that the follow-up n-th+m of theme time measurement period analysis draws is all its previous measurement period, wherein n > 1, m > 1.It converts theme subsequent association theme and refers to convert the relating subject of theme and convert the relating subject of the follow-up same subject of theme.
It can be seen that the present invention is based on the information gathering of internet information, news information cluster, information retrieval from above-mentioned embodiment, to the media event in known the Internet, calculate and analyze the related news theme of media event, related news report reprinting amount and related commentary number thereof.The distribution trend that media event develops in time is shown by theme and relevant report number change；And by thematic relation, reprinting amount and comment number analysis hot news theme in distribution trend.
Obviously, those skilled in the art can carry out various change and modification without departing from the spirit and scope of the present invention to the present invention.So, if these amendments of the present invention and modification belong within the scope of the claims in the present invention and equivalent technology thereof, then the present invention is also intended to comprise these change and modification.
Claims (13)
1. an internet news event report trend analysis, comprises the following steps:
(1) according to the characteristic information of the media event configured, the news information in Real-time Collection the Internet；
(2) news information described in pretreatment, filters out issuing time news information in the measurement period of the media event arranged；
(3) text message of the news information that Analysis and Screening goes out, obtains theme and the subject information of news information；Described theme includes the theme that there is subject information and there is not the source theme of subject information；
(4) analyze the theme in current statistic cycle and previous measurement period, draw the related information number of the same subject between current statistic cycle and previous measurement period and each same subject；
(5) according to described related information number, conversion theme is drawn；Described conversion theme refers to the related information number theme more than or equal to the related information number of previous this same subject of measurement period of this same subject of current statistic cycle；
(6) obtain the source theme that there is not subject information set in the time period and convert the related information number of theme, and representing and there is not the source theme of subject information and convert the related information number of theme distribution trend in time.
2. internet news event report trend analysis as claimed in claim 1, it is characterised in that: in step (1), described characteristic information includes the logical relation between the key word of news information, keyword weight and key word；Described news information includes text message and review information；Described keyword weight refers to key word status in overall crucial phrase；Logical relation between described key word include and or, non-and between nest relation, expressed by logical relation expression formula.
3. internet news event report trend analysis as claimed in claim 1, it is characterised in that: in step (1), in described collection the Internet, news information includes the download to described news information, analysis, the extraction of metadata.
4. internet news event report trend analysis as claimed in claim 3, it is characterized in that: described metadata includes that text message metadata and review information metadata, described text message metadata include title, text, the web site name in source, channel designation, issuing time；Described review information metadata includes commenting on text, comment issuing time and comment author.
5. the internet news event report trend analysis as according to any one of Claims 1 to 4, it is characterized in that: in step (2), the concrete grammar of described screening issuing time news information in the measurement period of set media event is as follows:
The news information gathered in a measurement period that () screening issuing time is media event；
B () checks whether the news information filtered out in step (a) exists subject information, in this way, then screened, and otherwise, filters qualified news information by the logical relation between the key word set, key word.
6. internet news event report trend analysis as claimed in claim 5, it is characterized in that: in step (b), when logical relation between the described key word by setting, key word filters qualified news information, only include body information, do not include news analysis information.
7. internet news event report trend analysis as claimed in claim 1, it is characterised in that: in step (3), theme is sorted from high to low by degree of association, only retain the N number of source theme forward by ranking in described relevancy ranking, N >=10.
8. internet news event report trend analysis as claimed in claim 1, it is characterized in that: in step (5), after drawing conversion theme, being sorted from high to low by the similarity with theme by the related information converted in theme, the information that labelling similarity is the highest represents article as it.
9. internet news event report trend analysis as claimed in claim 1, it is characterized in that: in step (6), represent source theme by scatterplot and convert the related information number distribution trend in time of theme, X-axis represents the time of theme, and Y-axis represents the related information number of theme.
10. the internet news event report trend analysis as described in claim 1 or 9, it is characterized in that: in step (6), the described related information number representing source theme distribution trend in time, represents including to source theme, the follow-up same subject of source theme, source theme relating subject and source theme subsequent association theme；
The described source follow-up same subject of theme refers to that the theme that n-th measurement period draws is all the same subject of its previous measurement period, wherein n > 1；Described source theme relating subject refers to the theme identical with previous measurement period；Described source theme subsequent association theme refers to relating subject and the relating subject of the follow-up same subject of source theme of source theme.
The 11. internet news event report trend analysis as described in claim 1 or 9, it is characterized in that: described in represent the related information number distribution trend in time converting theme, including to converting theme, convert the follow-up same subject of theme, convert representing of theme subsequent association theme；
The described follow-up same subject of conversion theme refers to assume that pth time measurement period analysis show that conversion theme, the theme that thus theme follow-up pth+q measurement period analysis of conversion draws are all the same subjects of its previous measurement period, wherein p > 1, q > 1；Described conversion theme subsequent association theme refers to convert the relating subject of theme and convert the relating subject of the follow-up same subject of theme.
12. 1 kinds of internet news event report trend analysis systems, including the information collecting device (11) of news information in the characteristic information Real-time Collection the Internet according to media event；
For filtering out the data prediction device (12) of issuing time news information in the measurement period of the media event arranged；
For the text message of the news information that Analysis and Screening goes out, obtain the theme of news information and the topic detection device (13) of subject information；Described theme includes the theme that there is subject information and there is not the source theme of subject information；
For analyzing and draw the thematic relation analytical equipment (14) of related information number of the same subject between current statistic cycle and previous measurement period and each same subject；
For according to the related information number of same subject between current statistic cycle and previous measurement period, drawing conversion subject analysis device (15) converting theme；
The source theme that there is not subject information in obtaining and representing the setting time period and the distribution trend demonstration device (16) of the related information number converting theme distribution trend in time.
13. internet news event as claimed in claim 12 report trend analysis systems, it is characterised in that: described data prediction device (12) includes that data Preliminary screening module (121) and data screen module (122) again；The news information that described data Preliminary screening module (121) are gathered in screening the measurement period that issuing time is media event；Described data screen module (122) again for when the news information of data Preliminary screening module screening does not exists subject information, screening news information further according to characteristic information.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210004878.3A CN103198078B (en) | 2012-01-09 | 2012-01-09 | A kind of internet news event report trend analysis and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201210004878.3A CN103198078B (en) | 2012-01-09 | 2012-01-09 | A kind of internet news event report trend analysis and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103198078A CN103198078A (en) | 2013-07-10 |
| CN103198078B true CN103198078B (en) | 2016-08-03 |
Family
ID=48720647
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201210004878.3A Expired - Fee Related CN103198078B (en) | 2012-01-09 | 2012-01-09 | A kind of internet news event report trend analysis and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103198078B (en) |
Families Citing this family (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9397904B2 (en) * | 2013-12-30 | 2016-07-19 | International Business Machines Corporation | System for identifying, monitoring and ranking incidents from social media |
| CN104268297A (en) * | 2014-10-28 | 2015-01-07 | 江苏惠居乐信息科技有限公司 | Big data analysis system on basis of news |
| CN105373579B (en) * | 2015-08-18 | 2018-08-03 | 天津大学 | A kind of news competitiveness analysis method and its visualization device based on regression analysis |
| CN108763429B (en) * | 2018-05-24 | 2022-03-29 | 成都睿码科技有限责任公司 | Two-dimensional display method for simulating development trend of cell-breeding news events |
| CN110598151B (en) * | 2019-09-09 | 2023-07-14 | 河南牧业经济学院 | Method and system for judging news spreading effect |
| CN113159363B (en) * | 2020-12-30 | 2022-04-19 | 成都信息工程大学 | Event trend prediction method based on historical news reports |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101231640A (en) * | 2007-01-22 | 2008-07-30 | 北大方正集团有限公司 | Method and system for automatically computing subject evolution trend in the internet |
| WO2010048430A3 (en) * | 2008-10-22 | 2010-07-22 | Fwix, Inc. | System and method for identifying trends in web feeds collected from various content servers |
| CN102073683A (en) * | 2010-12-22 | 2011-05-25 | 四川大学 | Distributed real-time news information acquisition system |
-
2012
- 2012-01-09 CN CN201210004878.3A patent/CN103198078B/en not_active Expired - Fee Related
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101231640A (en) * | 2007-01-22 | 2008-07-30 | 北大方正集团有限公司 | Method and system for automatically computing subject evolution trend in the internet |
| WO2010048430A3 (en) * | 2008-10-22 | 2010-07-22 | Fwix, Inc. | System and method for identifying trends in web feeds collected from various content servers |
| CN102073683A (en) * | 2010-12-22 | 2011-05-25 | 四川大学 | Distributed real-time news information acquisition system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103198078A (en) | 2013-07-10 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103198078B (en) | A kind of internet news event report trend analysis and system | |
| CN102760138B (en) | Classification method and device for user network behaviors and search method and device for user network behaviors | |
| Blismas et al. | Computer-aided qualitative data analysis: panacea or paradox? | |
| CN103164427B (en) | News Aggreagation method and device | |
| KR101463974B1 (en) | Big data analysis system for marketing and method thereof | |
| CN103365924B (en) | A kind of method of internet information search, device and terminal | |
| CN102567494B (en) | Website classification method and device | |
| CN103023714B (en) | The liveness of topic Network Based and cluster topology analytical system and method | |
| CN103246644B (en) | Method and device for processing Internet public opinion information | |
| CN105718587A (en) | Network content resource evaluation method and evaluation system | |
| CN101751458A (en) | Network public sentiment monitoring system and method | |
| CN103838785A (en) | Vertical search engine in patent field | |
| CN103020159A (en) | Method and device for news presentation facing events | |
| Weiler et al. | Event identification and tracking in social media streaming data | |
| CN103235796A (en) | Searching method and system based on clicking behavior of user | |
| US20150269138A1 (en) | Publication Scope Visualization and Analysis | |
| CN104391978A (en) | Method and device for storing and processing web pages of browsers | |
| CN103440328A (en) | User classification method based on mouse behaviors | |
| Yu et al. | Web content information extraction based on DOM tree and statistical information | |
| CN104484367A (en) | Data mining and analyzing system | |
| CN104376066A (en) | Network specific content digging method and device and electronic equipment | |
| CN109033133A (en) | Event detection and tracking based on Feature item weighting growth trend | |
| Xu et al. | Research on Tibetan hot words, sensitive words tracking and public opinion classification | |
| CN107463570B (en) | Document retrieval/analysis method and device | |
| Muthmann et al. | Detecting near-duplicate relations in user generated forum content |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20160803 Termination date: 20180109 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |