CN108170671A - A kind of method for extracting media event time of origin - Google Patents
A kind of method for extracting media event time of origin Download PDFInfo
- Publication number
- CN108170671A CN108170671A CN201711378174.1A CN201711378174A CN108170671A CN 108170671 A CN108170671 A CN 108170671A CN 201711378174 A CN201711378174 A CN 201711378174A CN 108170671 A CN108170671 A CN 108170671A
- Authority
- CN
- China
- Prior art keywords
- word
- news report
- article
- news
- report article
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 28
- 238000012545 processing Methods 0.000 claims abstract description 8
- 239000000284 extract Substances 0.000 claims description 5
- 239000000203 mixture Substances 0.000 claims description 3
- 238000000605 extraction Methods 0.000 description 8
- 230000002123 temporal effect Effects 0.000 description 7
- 238000011160 research Methods 0.000 description 5
- 238000005516 engineering process Methods 0.000 description 4
- 238000010586 diagram Methods 0.000 description 3
- 230000007812 deficiency Effects 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 238000012804 iterative process Methods 0.000 description 2
- 238000013456 study Methods 0.000 description 2
- 238000012549 training Methods 0.000 description 2
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 239000002360 explosive Substances 0.000 description 1
- 238000005065 mining Methods 0.000 description 1
- 238000003058 natural language processing Methods 0.000 description 1
- 238000002360 preparation method Methods 0.000 description 1
- 230000002035 prolonged effect Effects 0.000 description 1
- 238000012827 research and development Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/279—Recognition of textual entities
- G06F40/284—Lexical analysis, e.g. tokenisation or collocates
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
- G06F16/33—Querying
- G06F16/332—Query formulation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/279—Recognition of textual entities
- G06F40/289—Phrasal analysis, e.g. finite state techniques or chunking
- G06F40/295—Named entity recognition
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Computational Linguistics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Audiology, Speech & Language Pathology (AREA)
- General Health & Medical Sciences (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Mathematical Physics (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The embodiment of the invention discloses a kind of method for extracting media event time of origin, wherein, this method includes：News report article is obtained, each word in the news report article is extracted, is calculated with the title of news report article, obtain the degree of correlation between the two；The word in the news report article is obtained, is calculated, obtains the position of word distribution；To the word in the news report article by identifying processing, the keyword in the news report article is obtained；Keyword tissue in the news report article summarizes the main contents of entire chapter news report article, and the time of origin node of the news report article is obtained using search engine.Implement the embodiment of the present invention, obtain so as to extract that media event time of origin is more intelligent, accuracy rate is improved；The operation is more convenient, and greatly reduces job costs.
Description
Technical field
The present invention relates to information retrieval, text mining, natural language processing, field of artificial intelligence more particularly to one
The method of kind extraction media event time of origin.
Background technology
With the rapid development of Internet technology, information and information are brought all in data application field to people’s lives
It is mostly convenient, in the case that information data is in explosive increase, user's fast browsing in the data of magnanimity is helped to want what is obtained
Information is particularly important.News is exactly message, reflects the event occurred all the time, each media event having time attribute,
One of the characteristics of real-time is the maximum of news.According to the expression characteristic of news, news time of origin be media event can not or
How a scarce part, extract relevant temporal information from media event, it is referred to as most important the problem of.
Temporal information extraction belongs to the research field of information extraction, by prolonged research and development, to this problem
Solution achieves certain development：Research is expanded in terms of temporal information expression；Serial achievement is obtained in terms of timestamp
's；Research has been carried out in terms of time-event relation.However, the temporal information about media event is extracted, at present only
Research have focused largely on news report time and news briefing time, how accurately to extract the time of origin of media event,
Research both domestic and external is also fewer.
In the prior art, have and method is stabbed between to analyze the technology of news, which passes through the time in news report
The time of origin for stabbing to analyze and derive media event, which, which implements, mainly the defects of following two aspects：
First, complicated for operation, time word needs are formulated specific rule or template and can be just mapped in calendar date in this method, and
And each temporal information is done and is extracted based on covering downwards, heavy workload is not easy to operate；Second, for a news report,
Relevant timestamp information can not be found in most cases or find the timestamp information of great quantities of spare, so being difficult to effectively
Realize the extraction of media event time of origin.
Either based on URL time formats and the direct extracting time information of headline, this method is too simple, merely
Information is obtained by headline, frequently can lead to bit error rate height；Structure and linking URL time format library, job costs increase,
Not only cost higher, but also be difficult to accurately extract the real time of origin of media event.
Invention content
It is an object of the invention to overcome the deficiencies in the prior art, and the present invention provides during a kind of extraction media event generation
Between method, accurately and effectively obtain the time of origin of magnanimity news messages, and overcome existing method in accuracy and extension
The deficiency of property etc..
To solve the above-mentioned problems, the present invention proposes a kind of method for extracting media event time of origin, the method
Including：
News report article is obtained, each word in the news report article is extracted, with news report article
Title calculated, obtain the degree of correlation between the two；
The word in the news report article is obtained, is calculated, obtains the position of word distribution；
To the word in the news report article by identifying processing, the key in the news report article is obtained
Word；
Keyword tissue in the news report article summarizes the main contents of entire chapter news report article, utilizes
Search engine obtains the time of origin node of the news report article.
Preferably, the step of title with the news report article is calculated is mainly used to judge the word
Importance degree, the word of the degree of correlation of word-headline in word w and the title title of the news report article
Coincidence number represent.Wherein, i-th of word is w in set of words Wi, according to word wiWith m news report article dm's
The degree of correlation of headline obtains additional score accordinglyIts publicity is expressed as：
According to word and the matching result of headline, give word and add corresponding score, such as word wiIn headline
There is 1 word correlation, then word wiAdditional scoreEqual to 1, word wiIt is related to there is 2 words in headline,
Then word wiAdditional scoreEqual to 2, and so on.
Preferably, the step of word in the acquisition news report article is calculated mainly needs to consider word
The position occurred in article judges the importance of word, then sets m news report article dmHeadline in, n dm
In include the sum of word, according to word wiAppear in dmArticle position, obtain additional score accordingly
Mathematic(al) representation it is as follows：
Wherein, s represents the sentence of news report document, sjThe jth word of the news report is represented, for example, when setting threshold
During value j≤3, then it represents that first three sentence of news report is important content distribution region, wiIt appears within the region, phase can be obtained
The additional score answered.
Preferably, the word in the news report article is by identifying processing, including：
To name, place name, institution term, time, number in article etc. name group of entities into name entity sets NE,
If word wiBelong to name entity sets NE, then can obtain an additional fractional value accordingly,For document dmIn
Word wiThe additional score obtained about name Entity recognition.M news report document d in collection of document Dm, it is assumed that
Document dmIn each word wiFeature it is mutual indepedent, i-th words of some theme z in set of words is set as zi；P is closes
The probability that keyword occurs, each word correspond to the score SS obtained by this news reportmiIt is the linear combination of each characteristic component,
It is shown below：
Wherein, the weights omega of r ∈ { title, loc, ner }, factor alpha and each characteristic componentjSeek most on data set
Excellent combination；P is probability, to word according to score SSmiIt sorts from high to low, chooses n word composition keyword before score sequence
Collect Q, as the keyword in the news report article.
Preferably, the keyword tissue in the news report article summarizes the master of entire chapter news report article
Content is wanted, the time of origin node that the news report article is obtained using search engine is specifically included：
Based on the keyword acquired in LDA language models, each news dmThere is a corresponding keyword set Qm={ w1,
w2,w3,...,wn, summarize the main contents of a news report using several keywords in set；
Using the media event knowledge base that news search engine is perfect, a news search engine interface is connected, input is new
News term is Qm={ w1,w2,w3,...,wn, such as " the http under search engine Google news engine interfaces://
Google.search_news (query=Qm, num, start, country_code) ", input with media event is relevant looks into
Ask sentence QmIt is retrieved, a query statement is the query statement for representing media event, such as query statement " Trump
Speech ", and return to several news documents is crawled from several news documents of return using analytical tool before being ordered as
U news documents and corresponding issuing time respectively constitute initial return news list X={ x1,x2,x3,...,xUAnd it is corresponding
Issuing time sequence T={ t1,t2,t3,...tU, xiIt represents initial and returns to i-th of news documents in news list X, tiIt represents
In issuing time sequence T with i-th of news documents xiCorresponding issuing time, 1≤i≤U；
To each timing node t in the issuing time sequence T that parsesiFrequency fiIt is counted, obtains frequency most
High timing node, such as tiAppearance frequency fiIt is the maximum value in issuing time sequence T, then tiIt is then media event dm's
Time of origin.
In embodiments of the present invention, by constantly training, study and iterative process, news content keyword side is being extracted
Face, it is more intelligent, and accuracy rate greatly improves.Method by connecting news knowledge base is returned new by search engine
The method for hearing temporal information obtains the initial time of origin of media event, implement it is easier, greatly reduce work into
This.
Description of the drawings
In order to illustrate more clearly about the embodiment of the present invention or technical scheme of the prior art, to embodiment or will show below
There is attached drawing needed in technology description to be briefly described, it should be apparent that, the accompanying drawings in the following description is only this
Some embodiments of invention, for those of ordinary skill in the art, without creative efforts, can be with
Other attached drawings are obtained according to these attached drawings.
Fig. 1 is the method flow schematic diagram of the extraction news time time of origin of the embodiment of the present invention；
Fig. 2 is the time of origin node for obtaining the news report article in the embodiment of the present invention using search engine
Flow diagram.
Specific embodiment
Below in conjunction with the attached drawing in the embodiment of the present invention, the technical solution in the embodiment of the present invention is carried out clear, complete
Site preparation describes, it is clear that described embodiment is only part of the embodiment of the present invention, instead of all the embodiments.It is based on
Embodiment in the present invention, those of ordinary skill in the art are obtained every other without creative efforts
Embodiment shall fall within the protection scope of the present invention.
Fig. 1 is a kind of method flow schematic diagram of extraction media event time of origin of the embodiment of the present invention, such as Fig. 1 institutes
Show, this method includes：
S1 obtains news report article, each word in the news report article is extracted, with the Xin Wen Bao
The title of road article is calculated, and obtains the degree of correlation between the two；
S2, the word obtained in the news report article are calculated, and obtain the position of word distribution；
S3 to the word in the news report article by identifying processing, obtains the pass in the news report article
Keyword；
S4, the keyword tissue in the news report article summarize the main contents of entire chapter news report article,
The time of origin node of the news report article is obtained using search engine.
The embodiment of the present invention is to employ bayesian probability theory to extract keyword to news report article.Its
In, we carry out establishing LDA language models based on bayesian probability theory, and the model includes three layers, and master is obtained from article
Topic, extracts some theme from theme branch, extracts a word from the word distribution corresponding to the theme, repeats
The above process finally obtains the keyword in the article until each word of traversal.
Specifically, being calculated with the title of the news report article described in S1, is mainly used to judge the word
Importance degree, the word of the degree of correlation of word-headline in word w and the title title of the news report article
Coincidence number represent.Wherein, i-th of word is w in set of words Wi, according to word wiWith m news report article dm's
The degree of correlation of headline obtains additional score accordinglyIts formula is expressed as：
According to word and the matching result of headline, give word and add corresponding score, such as word wiWith it is new
Hearing in title has 1 word correlation, then word wiAdditional scoreEqual to 1, word wiAnd have 2 words in headline
Language is related, then word wiAdditional scoreEqual to 2, and so on.
S2 is described further：
Method based on word position is that field is relevant, such as in some fields, and a word of paragraph includes theme
Information, and some fields then appear in last sentence.In news report, the high sentence of information content typically occurs in former sentences
With section head, the position that consideration word occurs in article is needed to judge the importance of word, then sets m news report articles
dmHeadline in, n dmIn include the sum of word, according to word wiAppear in dmArticle position, obtain corresponding attached
Bonus point numberThen its formula is expressed as：
Wherein, s represents the sentence of news report document, slThe l words of the news report are represented, for example, when setting threshold
During value l≤3, then it represents that first three sentence of news report is important content distribution region, as word wiIt appears within the region, it can be with
Obtain additional score accordingly.
Wherein, described in S3 by identifying processing it is to employ name entity to the word in the news report article
Identifying processing names entity to name, place name, institution term, time, number in article etc..
S3 is described further：
To name, place name, institution term, time, number in article etc. name group of entities into name entity sets NE,
If word wiBelong to name entity sets NE, then can obtain an additional fractional value accordingly,For document dmIn
Word wiThe additional score obtained about name Entity recognition.M news report document d in collection of document Dm, it is assumed that
Document dmIn each word wiFeature it is mutual indepedent, i-th words of some theme z in set of words is set as zi；P is closes
The probability that keyword occurs, each word correspond to the score SS obtained by this news reportmiIt is the linear combination of each characteristic component,
It is shown below：
Wherein, the weights omega of r ∈ { title, loc, ner }, factor alpha and each characteristic componentjSeek most on data set
Excellent combination；P is probability, to word according to score SSmiIt sorts from high to low, chooses n word composition keyword before score sequence
Collect Q, as the keyword in the news report article.
S4 is described further：
Based on the keyword acquired in LDA language models, each news dmThere is a corresponding keyword set Qm={ w1,
w2,w3,...,wn, summarize the main contents of a news report using several keywords in set；
As shown in Fig. 2, using the perfect media event knowledge base of news search engine, one news search engine of connection connects
Mouthful, input news retrieval word is Qm={ w1,w2,w3,...,wn, such as under search engine Google news engine interfaces
“http://Google.search_news (query=Qm, num, start, country_code) ", input and media event
Relevant query statement QmIt is retrieved, a query statement is the query statement such as query statement " Trump for representing media event
Speech ", and return to several news documents is crawled from several news documents of return using analytical tool before being ordered as
U news documents and corresponding issuing time respectively constitute initial return news list X={ x1,x2,x3,...,xUAnd it is corresponding
Issuing time sequence T={ t1,t2,t3,...tU, xiIt represents initial and returns to i-th of news documents in news list X, tiIt represents
In issuing time sequence T with i-th of news documents xiCorresponding issuing time, 1≤i≤U；
To each timing node t in the issuing time sequence T that parsesiFrequency fiIt is counted, obtains frequency most
High timing node, such as tiAppearance frequency fiIt is the maximum value in issuing time sequence T, then tiIt is then media event dm's
Time of origin.
In embodiments of the present invention, by constantly training, study and iterative process, news content keyword side is being extracted
Face, it is more intelligent, and accuracy rate greatly improves.Method by connecting news knowledge base is returned new by search engine
The method for hearing temporal information obtains the initial time of origin of media event, implement it is easier, greatly reduce work into
This.
One of ordinary skill in the art will appreciate that all or part of step in the various methods of above-described embodiment is can
It is completed with instructing relevant hardware by program, which can be stored in a computer readable storage medium, storage
Medium can include：Read-only memory (ROM, Read Only Memory), random access memory (RAM, Random
Access Memory), disk or CD etc..
In addition, a kind of method of extraction media event time of origin provided above the embodiment of the present invention has carried out in detail
Thin to introduce, specific case used herein is expounded the principle of the present invention and embodiment, and above example is said
The bright method and its core concept for being merely used to help understand the present invention；Meanwhile for those of ordinary skill in the art, foundation
The thought of the present invention, there will be changes in specific embodiments and applications, in conclusion the content of the present specification is not
It is interpreted as limitation of the present invention.
Claims (4)
- A kind of 1. method for extracting media event time of origin, which is characterized in that the method includes：News report article is obtained, extracts each word in the news report article, the mark with news report article Topic is calculated, and obtains the degree of correlation between the two；The word in the news report article is obtained, is calculated, obtains the position of word distribution；To the word in the news report article by identifying processing, the keyword in the news report article is obtained；Keyword tissue in the news report article summarizes the main contents of entire chapter news report article, utilizes search Engine obtains the time of origin node of the news report article.
- 2. a kind of method for extracting media event time of origin as described in claim 1, which is characterized in that it is described with it is described The step of title of news report article is calculated is mainly used to judge the importance degree of the word, word-headline The degree of correlation represented with word w with the number that overlaps of the word in the title title of the news report article.Wherein, word collection It is w to close i-th of word in Wi, according to word wiWith m news report article dmHeadline the degree of correlation, obtain corresponding Additional scoreIts publicity is expressed as：According to word and headline Matching result, give word and add corresponding score, such as word wiIt is related to there is 1 word in headline, then word Language wiAdditional scoreEqual to 1, word wiIt is related to there is 2 words in headline, then word wiAdditional scoreEqual to 2, and so on.
- A kind of 3. method for extracting media event time of origin as described in claim 1, which is characterized in that the acquisition institute Stating the step of word in news report article is calculated mainly needs to consider that grammatical term for the character is carried out in the position that word occurs in article The importance of language then sets m news report article dmHeadline in, n dmIn include the sum of word, according to word wiAppear in dmArticle position, obtain additional score accordinglyThe following institute of mathematic(al) representation Show：Wherein, s represents the sentence of news report document, sjThe jth word of the news report is represented, for example, when setting threshold value j≤3 When, then it represents that news report first three sentence is important content distribution region, wiIt appears within the region, can obtain corresponding attached Bonus point number.
- 4. a kind of method for extracting media event time of origin as described in claim 1, which is characterized in that described to described new The word in report article is heard by identifying processing, including：If word wiBelong to name entity sets NE, then can obtain an additional fractional value accordingly,For document dmIn word wiThe additional score obtained about name Entity recognition.M news report document d in collection of document Dm, Assuming that document dmIn each word wiFeature it is mutual indepedent, i-th words of some theme z in set of words is set as zi；Its In, p is the probability that keyword occurs, and each word corresponds to the score SS obtained by this news reportmiIt is each characteristic component Linear combination is shown below：Wherein, the weights omega of r ∈ { title, loc, ner }, factor alpha and each characteristic componentjSeek optimal set on data set It closes；P is probability, to word according to score SSmiIt sorts from high to low, chooses n word composition keyword set Q before score sequence, As the keyword in the news report article.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711378174.1A CN108170671A (en) | 2017-12-19 | 2017-12-19 | A kind of method for extracting media event time of origin |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201711378174.1A CN108170671A (en) | 2017-12-19 | 2017-12-19 | A kind of method for extracting media event time of origin |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108170671A true CN108170671A (en) | 2018-06-15 |
Family
ID=62522626
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201711378174.1A Pending CN108170671A (en) | 2017-12-19 | 2017-12-19 | A kind of method for extracting media event time of origin |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108170671A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110232183A (en) * | 2018-12-07 | 2019-09-13 | 腾讯科技（深圳）有限公司 | Keyword extraction model training method, keyword extracting method, device and storage medium |
| CN111191413A (en) * | 2019-12-30 | 2020-05-22 | 北京航空航天大学 | Method, device and system for automatically marking event core content based on graph sequencing model |
| CN111814770A (en) * | 2020-09-04 | 2020-10-23 | 中山大学深圳研究院 | Content keyword extraction method of news video, terminal device and medium |
| CN112188312A (en) * | 2019-07-02 | 2021-01-05 | 百度（美国）有限责任公司 | Method and apparatus for determining video material of news |
| CN112612944A (en) * | 2020-12-07 | 2021-04-06 | 深圳价值在线信息科技股份有限公司 | Case information management method, terminal equipment and system |
| CN113536777A (en) * | 2021-07-30 | 2021-10-22 | 深圳豹耳科技有限公司 | Extraction method, device and equipment of news keywords and storage medium |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104182504A (en) * | 2014-08-18 | 2014-12-03 | 合肥工业大学 | Algorithm for dynamically tracking and summarizing news events |
| CN104408093A (en) * | 2014-11-14 | 2015-03-11 | 中国科学院计算技术研究所 | News event element extracting method and device |
| CN104715014A (en) * | 2015-01-26 | 2015-06-17 | 中山大学 | Online news topic detection method |
| CN104915446A (en) * | 2015-06-29 | 2015-09-16 | 华南理工大学 | Automatic extracting method and system of event evolving relationship based on news |
| CN105354333A (en) * | 2015-12-07 | 2016-02-24 | 天云融创数据科技（北京）有限公司 | Topic extraction method based on news text |
| CN106599285A (en) * | 2016-12-23 | 2017-04-26 | 北京奇虎科技有限公司 | News searching-based searching result providing method and apparatus |
-
2017
- 2017-12-19 CN CN201711378174.1A patent/CN108170671A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104182504A (en) * | 2014-08-18 | 2014-12-03 | 合肥工业大学 | Algorithm for dynamically tracking and summarizing news events |
| CN104408093A (en) * | 2014-11-14 | 2015-03-11 | 中国科学院计算技术研究所 | News event element extracting method and device |
| CN104715014A (en) * | 2015-01-26 | 2015-06-17 | 中山大学 | Online news topic detection method |
| CN104915446A (en) * | 2015-06-29 | 2015-09-16 | 华南理工大学 | Automatic extracting method and system of event evolving relationship based on news |
| CN105354333A (en) * | 2015-12-07 | 2016-02-24 | 天云融创数据科技（北京）有限公司 | Topic extraction method based on news text |
| CN106599285A (en) * | 2016-12-23 | 2017-04-26 | 北京奇虎科技有限公司 | News searching-based searching result providing method and apparatus |
Non-Patent Citations (1)
| Title |
|---|
| 王升翔等: "基于分层树模型的中文网页主题时间提取方法", 《计算机应用》 * |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110232183A (en) * | 2018-12-07 | 2019-09-13 | 腾讯科技（深圳）有限公司 | Keyword extraction model training method, keyword extracting method, device and storage medium |
| CN110232183B (en) * | 2018-12-07 | 2022-05-27 | 腾讯科技（深圳）有限公司 | Keyword extraction model training method, keyword extraction device and storage medium |
| US11947911B2 (en) | 2018-12-07 | 2024-04-02 | Tencent Technology (Shenzhen) Company Limited | Method for training keyword extraction model, keyword extraction method, and computer device |
| CN112188312A (en) * | 2019-07-02 | 2021-01-05 | 百度（美国）有限责任公司 | Method and apparatus for determining video material of news |
| CN112188312B (en) * | 2019-07-02 | 2023-10-27 | 百度（美国）有限责任公司 | Method and device for determining video material of news |
| CN111191413A (en) * | 2019-12-30 | 2020-05-22 | 北京航空航天大学 | Method, device and system for automatically marking event core content based on graph sequencing model |
| CN111191413B (en) * | 2019-12-30 | 2021-11-12 | 北京航空航天大学 | Method, device and system for automatically marking event core content based on graph sequencing model |
| CN111814770A (en) * | 2020-09-04 | 2020-10-23 | 中山大学深圳研究院 | Content keyword extraction method of news video, terminal device and medium |
| CN112612944A (en) * | 2020-12-07 | 2021-04-06 | 深圳价值在线信息科技股份有限公司 | Case information management method, terminal equipment and system |
| CN112612944B (en) * | 2020-12-07 | 2024-05-31 | 深圳价值在线信息科技股份有限公司 | Case information management method, terminal equipment and system |
| CN113536777A (en) * | 2021-07-30 | 2021-10-22 | 深圳豹耳科技有限公司 | Extraction method, device and equipment of news keywords and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108170671A (en) | A kind of method for extracting media event time of origin | |
| CN111753099B (en) | Method and system for enhancing relevance of archive entity based on knowledge graph | |
| CN102955856B (en) | Chinese short text classification method based on characteristic extension | |
| CN111143479A (en) | Knowledge graph relation extraction and REST service visualization fusion method based on DBSCAN clustering algorithm | |
| KR101646754B1 (en) | Apparatus and Method of Mobile Semantic Search | |
| CN106095762A (en) | A kind of news based on ontology model storehouse recommends method and device | |
| CN111190900B (en) | JSON data visualization optimization method in cloud computing mode | |
| CN104021198B (en) | The relational database information search method and device indexed based on Ontology | |
| CN103455487B (en) | The extracting method and device of a kind of search term | |
| CN105468605A (en) | Entity information map generation method and device | |
| CN109597895B (en) | Knowledge graph-based official document searching method | |
| JP2005526317A (en) | Method and system for automatically searching a concept hierarchy from a document corpus | |
| CN107436955B (en) | English word correlation degree calculation method and device based on Wikipedia concept vector | |
| CN101393565A (en) | Facing virtual museum searching method based on noumenon | |
| CN111177591A (en) | Knowledge graph-based Web data optimization method facing visualization demand | |
| CN101004762A (en) | Network web page system of a dynamic multidimensional Internet | |
| Agosti et al. | On the use of information retrieval techniques for the automatic construction of hypertext | |
| CN105677638B (en) | Web information abstracting method | |
| CN102810114A (en) | Personal computer resource management system based on body | |
| CN107656921B (en) | Short text dependency analysis method based on deep learning | |
| CN105608232A (en) | Bug knowledge modeling method based on graphic database | |
| CN107330111A (en) | The search method and device of domain body based on common version body | |
| CN107977420A (en) | The abstract extraction method, apparatus and readable storage medium storing program for executing of a kind of evolved document | |
| CN105260396A (en) | Word retrieval method and apparatus | |
| CN100562872C (en) | Automatic moulding plate information locating method at the structuring webpage |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180615 |