CN101957812A - Verb semantic information extracting method based on event ontology - Google Patents
Verb semantic information extracting method based on event ontology Download PDFInfo
- Publication number
- CN101957812A CN101957812A CN2010102908605A CN201010290860A CN101957812A CN 101957812 A CN101957812 A CN 101957812A CN 2010102908605 A CN2010102908605 A CN 2010102908605A CN 201010290860 A CN201010290860 A CN 201010290860A CN 101957812 A CN101957812 A CN 101957812A
- Authority
- CN
- China
- Prior art keywords
- verb
- role
- sentence
- information
- element array
- 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
Images
Landscapes
- Machine Translation (AREA)
Abstract
本发明涉及一种基于事件本体的动词语义信息提取方法。它是首先根据输入的句子得到句子要素数组A；其次利用角色提取规则对句子要素数组A中的动词角色进行提取；接着通过动词和动词角色匹配动词概念的方法确定动词词义；再接着根据时间时态提取规则识别动词的时间和时态信息；最后根据动词角色、动词词义和动词时间时态信息生成动词语义信息。由于本发明基于事件本体，通过匹配动词和动词角色的方法，提高了识别动词的准确率，由于通过动词词义信息、时间时态信息及动词与动词角色间关系信息，生成了丰富的动词语义信息，从而解决了已有技术存在的动词识别准确率低和动词语义表达不足的问题。
The invention relates to a method for extracting verb semantic information based on event ontology. It first obtains the sentence element array A according to the input sentence; secondly, uses the role extraction rule to extract the verb role in the sentence element array A; then determines the meaning of the verb by matching the verb concept with the verb and the verb role; and then according to the time The time and tense information of verbs are identified through tense extraction rules; finally, verb semantic information is generated according to verb roles, verb meanings and verb time tense information. Because the present invention is based on the event ontology, by matching verbs and verb roles, the accuracy of verb recognition is improved, and rich verb semantic information is generated through verb meaning information, time tense information and relationship information between verbs and verb roles , thereby solving the problems of low verb recognition accuracy and insufficient verb semantic expression existing in the prior art.
Description
Technical field
The invention belongs to natural language information and extract field, specifically a kind of semantic verbs information extracting method based on the incident body.
Background technology
Human development is to today, and a large amount of information is stored with the mode of electronic document, and nearly ten or twenty over year the quantity of these electronic documents present explosive growth.Information extraction technology is exactly to utilize computer technology to find from the electronic document of these magnanimity to meet the information that user interest requires.
Traditional information extracting method generally is divided into two kinds:
(1) rule-based approach.This method by the human rule of formulating, is mated information by computer technology in advance, and the message block that the match is successful is extracted.This method is understood by people more easily, and the less computer-chronograph execution speed of regular quantity is very fast.But it is bigger to work as regular quantity, the management between the rule, and the consistency detection between rule, it is difficult and complicated that cost computing time of rule all can become.
(2) based on the method for adding up.This method has good mathematical theory to support, and is the focus of the research of aspect natural language processing in recent years.It predicts the feature of unknown message from statistical angle by analyzing existing information characteristics.From the experiment effect of present research, it also is feasible carrying out information extraction with this method.But needing a large amount of corpus and complicated parameter to be provided with just, this method can obtain more satisfactory result.Because the corpus that uses based on the method for statistics can not comprise the whole field of natural language, cause the incomplete information extraction precision that has influenced based on statistical method of corpus.
More than rule-based and information is extracted based on the classic method of statistics, be that aspects such as the position of morphology, speech from text and part of speech are considered.But the extraction for semantic information but is helpless.
In recent years, the focus of another research has appearred in natural language processing field---and utilize ontology knowledge to carry out information extraction.This method utilizes body to make up the leaching process of domain knowledge tutorial message.The article report is for example arranged, its exercise question is: (this article author is a kind of improved Web information extraction based on body: Liu Jiagang, Chen Shan, Huang Ying, computer engineering) and article A Design of Temporal Event Extraction from Chinese Financial News (this article author Wenjie Li be published in the periodical of publishing in 2010:, Kam-Fai Wong, Chunfa Yuan is published in the periodical of publishing in 2003: International Journal of Computer Processing of Oriental Languages).These two pieces of articles disclose by notion, relation and rule in the body and have obtained certain effect as the foundation of information extraction.Yet this kind method also exists deficiency at present: priori in (1) shortage body and the relationship description between the information extraction; (2) recognition accuracy to verb in the sentence is still waiting to improve; (3) shortage is to the abundant expression of semantic verbs.Therefore, existing method is all not enough to some extent to the expression of the accuracy rate of verb identification in the sentence and semantic verbs.
Summary of the invention
The problem and shortage that exists of prior art in view of the above, the objective of the invention is to solve the problem that prior art exists, a kind of semantic verbs information extracting method based on the incident body is provided, this method has not only improved the accuracy rate of identification verb by coupling verb and verb role's method, and generates abundant semantic verbs information by the relation information between verb word sense information, time temporal information and verb and the verb role.
In order to reach above purpose, the present invention adopts following technical proposals:
A kind of semantic verbs information extracting method based on the incident body is characterized in that at first obtaining sentence key element array A according to the sentence of importing; Next utilizes role's extracting rule that verb role among the sentence key element array A is extracted; Then the method for mating the verb notion by verb and verb role is determined the verb meaning of a word; Then discern the time and the temporal information of verb again according to time tense extracting rule; Generate semantic verbs information according to verb role, the verb meaning of a word and verb time temporal information at last, its concrete steps comprise:
A, obtain sentence key element array A: the input sentence, from sentence, obtain the speech that meets the sentence key element, form sentence key element array A by these speech;
B, verb role extract: utilize the verb role among the verb role extracting rule extraction sentence key element array A;
C, the identification of the verb meaning of a word: the method by verb notion in verb and the verb role match event body is determined the verb meaning of a word;
D, the tense identification of verb time: according to the time and the temporal information of time tense extracting rule identification verb;
E, generation semantic verbs information: generate semantic verbs information according to the verb time temporal information of discerning among the verb meaning of a word of discerning among the verb role who extracts among the above-mentioned steps B, the above-mentioned steps C and the above-mentioned steps D.
Obtain sentence key element array A described in the above-mentioned steps A, its operation steps is as follows:
A1, use the participle instrument to carry out participle and word mark part of speech to cutting out to the sentence of input;
If do not have gerund or verb to ignore this sentence in the A2 sentence, promptly the processing of semantic verbs information extraction do not done in this sentence;
A3, according to participle in the steps A 1 and part-of-speech tagging result, the speech of the structural word, noun, gerund and the verb that meet BA-sentence that the sentence key element requires or " quilt " words and expressions in the sentence is extracted, and to be unit with the speech leave among the sentence key element array A by the sequencing of speech in former;
Verb role described in the above-mentioned steps B extracts, and is to utilize verb role extracting rule that verb role among the sentence key element array A is extracted, and its operating process is as follows:
According to noun composition among the sentence key element array A, noun composition architectural feature, utilize verb role extracting rule that the verb role among the sentence key element array A is extracted at verb front-back direction and " quilt " words and expressions and BA-sentence;
The identification of the meaning of a word of verb described in the above-mentioned steps C is to determine the verb meaning of a word by the method for verb notion in verb and the verb role match event body, and its operating process is as follows:
Verb and verb role among C1, the traversal sentence key element array A are mated, and its operating process is as follows:
Verb among the traversal sentence key element array A, in the incident body, inquire about and judge the character of this verb, if this verb is an intransitive verb, this verb only mates the preceding nearest verb role of verb, if this verb is a body guest verb, nearest verb role behind this preceding nearest verb role of verb coupling verb and the verb is if this verb is a meaning guest verb, nearest verb role before this verb coupling verb obtains verb and verb role and mates set M;
C2, judge that whether verb and verb role are mated set M is empty, if be empty, then abandon this extraction semantic verbs information processing, otherwise whether the incident body of utilization is judged verb and verb role and is mated, if do not match then change the part of speech of this verb into gerund, and again to the coupling of a last verb to discerning again, otherwise it is right to keep this coupling; Neither one mates remaining among the set M if final verb and verb role are mated, then abandon this extraction semantic verbs information processing, otherwise traversal verb and verb role are mated the element of gathering M, to verb notion in the mapping incident body, obtain the verb word sense information by every pair of verb and verb role's coupling;
The identification of the time temporal feature of verb described in the above-mentioned steps D, its operating process is as follows:
Extract the temporal information of sentence and tense adverbial word (as:,, speech such as back), the temporal information and the temporal information of identification verb according to time tense extracting rule;
Generate semantic verbs information described in the above-mentioned steps E, its operating process is as follows:
Extract according to step B that verb role, step C obtain the verb word sense information and step D obtains the semantic verbs information that the time temporal information generates sentence.
The present invention has following conspicuous outstanding substantive distinguishing features and marked improvement compared with prior art:
The present invention is based on the incident body,, improved the accuracy rate of identification verb by coupling verb and verb role's method; By relation information between verb word sense information, time temporal information and verb and verb role, generate abundant semantic verbs information.Solved the low and not enough problem of semantic verbs expression of verb recognition accuracy that prior art exists.
Description of drawings
Fig. 1 is the module rack composition based on the semantic verbs information extracting method of incident body that the present invention carries;
Fig. 2 is the overview flow chart based on the semantic verbs information extracting method of incident body that the present invention carries;
Fig. 3 is a verb role extracting rule structural representation in the mentioned verb role extraction steps of the present invention;
Fig. 4 is the process flow diagram of the identification verb meaning of a word in the mentioned verb meaning of a word identification step of the present invention;
Fig. 5 is the verb character decision flow chart of the mentioned judgement verb character of the present invention;
Fig. 6 is the synoptic diagram of this volume elements of adopting when making up of the mentioned incident body of the present invention.
Embodiment
Hereinafter a preferred embodiment that provides the semantic verbs information extracting method based on the incident body of the present invention according to Fig. 1 to 6 it is to be noted, given embodiment is technical characterstic and the functional characteristics that is used for illustrating the inventive method, enable to be easier to understand the present invention, rather than be used for limiting the scope of the invention.
With reference to Fig. 1, this is as follows based on the module architectures that the semantic verbs information extracting method of incident body comprises:
(1) obtain sentence key element array A201: the sentence to input uses the participle instrument to carry out participle and the word mark part of speech to cutting out, if do not have gerund or verb in the sentence, then ignore this sentence, otherwise the structural word of the BA-sentence in the sentence or " quilt " words and expressions, the speech of noun, gerund and verb extracts, and to be unit with the speech leave among the sentence key element array A by the sequencing of speech in former sentence;
(2) the verb role extracts 202: according to the noun composition among the sentence key element array A, the noun composition architectural feature at verb front-back direction and " quilt " words and expressions and BA-sentence, utilize role's extracting rule 205 to extract verb role among the candidate events sentence array A;
(3) semantic verbs identification 203: traversal sentence key element is formed the verb among the array A, in incident body 206, inquire about and judge the character of this verb, if this verb is an intransitive verb, this verb only mates the preceding nearest verb role of verb, if this verb is a body guest verb, nearest verb role behind this preceding nearest verb role of verb coupling verb and the verb is if this verb is a meaning guest verb, nearest verb role before this verb coupling verb obtains verb and verb role and mates set M; Judge whether verb and verb role are mated set M is empty, if be empty, then abandon this extraction semantic verbs information processing, otherwise whether the incident body of utilization is judged verb and verb role and is mated, if do not match then change the part of speech of this verb into gerund, and again to the coupling of a last verb to discerning again, otherwise it is right to keep this coupling; Neither one mates remaining among the set M if final verb and verb role are mated, then abandon this extraction semantic verbs information processing, otherwise traversal verb and verb role are mated the element of gathering M, to verb notion in the mapping incident body, obtain the verb word sense information by every pair of verb and verb role's coupling;
(4) the verb time sequence feature identification 204: according to time tense extracting rule 207 extracting time information and tense adverbial word (as:,, speech such as back), the time and the temporal information of identification verb;
(5) generate semantic verbs information 301: extract 202 according to the verb role and extract verb role, semantic verbs identification 203 and obtain verb word sense information and verb time sequence identification 204 and obtain the time temporal information and generate semantic verbs information in the sentence.
With reference to Fig. 2, this overview flow chart based on the semantic verbs information extracting method of incident body comprises the steps:
A, obtain sentence key element array A, the speech that meets the sentence key element obtained in the input sentence from sentence, form sentence key element array A by these speech;
B, verb role extract, and utilize the verb role among the verb role extracting rule extraction sentence key element array A;
C, the identification of the verb meaning of a word are determined the verb meaning of a word by the method for verb notion in verb and the verb role match event body;
D, the tense identification of verb time are according to the time and the temporal information of time tense extracting rule identification verb;
E, generation semantic verbs information generate semantic verbs information according to the verb time temporal information of discerning among the verb meaning of a word of discerning among the verb role who extracts among the above-mentioned steps B, the above-mentioned steps C and the above-mentioned steps D.
Obtain sentence key element array A described in the above-mentioned steps A, its operation steps is as follows:
A1, use the participle instrument to carry out participle and word mark part of speech to cutting out to the sentence of input;
If do not have gerund or verb to ignore this sentence in the A2 sentence, promptly the processing of semantic verbs information extraction do not done in this sentence;
A3, according to participle in the steps A 1 and part-of-speech tagging result, the speech of the structural word, noun, gerund and the verb that meet BA-sentence that the sentence key element requires or " quilt " words and expressions in the sentence is extracted, and to be unit with the speech leave among the sentence key element array A by the sequencing of speech in former;
Verb role described in the above-mentioned steps B extracts, and is to utilize verb role extracting rule that verb role among the sentence key element array A is extracted, and its operating process is as follows:
According to noun composition among the sentence key element array A, noun composition architectural feature, utilize verb role extracting rule that the verb role among the sentence key element array A is extracted at verb front-back direction and " quilt " words and expressions and BA-sentence;
The identification of the meaning of a word of verb described in the above-mentioned steps C is to determine the verb meaning of a word by the method for verb notion in verb and the verb role match event body, and its operating process is as follows:
Verb and verb role among C1, the traversal sentence key element array A are mated, and its operating process is as follows:
Verb among the traversal sentence key element array A, in the incident body, inquire about and judge the character of this verb, if this verb is an intransitive verb, this verb only mates the preceding nearest verb role of verb, if this verb is a body guest verb, nearest verb role behind this preceding nearest verb role of verb coupling verb and the verb is if this verb is a meaning guest verb, nearest verb role before this verb coupling verb obtains verb and verb role and mates set M;
C2, judge that whether verb and verb role are mated set M is empty, if be empty, then abandon this extraction semantic verbs information processing, otherwise whether the incident body of utilization is judged verb and verb role and is mated, if do not match then change the part of speech of this verb into gerund, and again to the coupling of a last verb to discerning again, otherwise it is right to keep this coupling; Neither one mates remaining among the set M if final verb and verb role are mated, then abandon this extraction semantic verbs information processing, otherwise traversal verb and verb role are mated the element of gathering M, to verb notion in the mapping incident body, obtain the verb word sense information by every pair of verb and verb role's coupling;
The identification of the time temporal feature of verb described in the above-mentioned steps D, its operating process is as follows:
Extract the temporal information of sentence and tense adverbial word (as:,, speech such as back), the temporal information and the temporal information of identification verb according to time tense extracting rule;
Generate semantic verbs information described in the above-mentioned steps E, its operating process is as follows:
Extract according to step B that verb role, step C obtain the verb word sense information and step D obtains the semantic verbs information that the time temporal information generates sentence.
Shown in Figure 3, verb role extracting rule most preferred embodiment is to be unit with sentence key element array A, extract a kind of template for the computing machine Automatic Extraction of verb role from sentence key element array A, it is the sequence that is made of role's speech of verb, verb, " quilt " words and expressions and BA-sentence feature and other statement compositions.
Such as: [{ * } { Actor}#{act_word}{*}]+
In verb role extracting rule, part in [] is the pattern of coupling, part in { } is a sentence element, * represent the sentence element except that verb role, verb and sentence elements such as " quilt " words and expressions and BA-sentence in the sentence, Actor represents role's speech of verb, act_word represents verb, and { } # represents that content can occur 0 time or 1 time in { }, and the content in []+expression [] occurs 1 time at least.
With reference to Fig. 4, as follows to this process step based on verb meaning of a word identification in the semantic verbs information extracting method of incident body:
(1) being arranged on the initial value that extracts i verb among the sentence key element array A is 1;
(2) in array A, obtain i verb;
(3) judge that whether the verb among the array A all travels through, if traveled through, then changes step (15), otherwise changes step (4);
(4) analyze the character of this verb,, then change step (5),, then change step (6), if this verb is meaning guest's verb character then changes step (7) if this verb is a body guest verb character if this verb is an intransitive verb character;
(5) mate the former piece role of role nearest before the verb, and this former piece role and verb, change step (8) as this verb;
(6) the nearest role before and after the verb respectively as the former piece role and the consequent role of this verb, and this former piece role and consequent role mated commentaries on classics step (8) with verb;
(7) mate the former piece role of role nearest before the verb, and this former piece role and verb as this verb;
(8) according to the restriction relation of verb that defines in the incident body and verb role coupling this verb and verb role are mated judging,, otherwise change step (9) if coupling is correctly changeed step (13);
(9) value with i subtracts 1;
(10) the value situation of judgement i: if i equals 0, then change step (11), otherwise change step (12);
(11) value that i is set is 1;
(12) change this verb part of speech into gerund, change step (2);
(13) value with i adds 1;
(14) will identify correct coupling and mate set M, change step (2) adding verb and verb role;
(15) judge whether set M is empty, if be empty, changes step (17), if be not empty, changes step (16);
(16) verb among the traversal set M and verb role are mated rightly, to shining upon verb notion in the incident body, obtain the verb word sense information by every pair of verb and verb role's coupling.
(17) finish.
With reference to Fig. 5, this is judged that based on verb character in the semantic verbs information extracting method of incident body process step is as follows:
Shown in Figure 6 is the synoptic diagram of this volume elements of adopting when making up of the mentioned incident body of present embodiment.Present embodiment is by OWL (Web Ontology Language) language description incident body, and this volume elements that relates in the incident Ontology Modeling has:
(1) notion of actConcept class 601 expression verbs;
(2) classification of eventClass class 602 presentation of events is as traffic hazard class, tsunami class etc.;
(3) the verb character type of ActoeProperty data attribute 603 expression verb notions 601, this verb character type is divided three classes: intransitive verb, body guest verb and meaning guest verb;
(4) the verb role of 604 expressions of MatchRole data attribute and verb notion 601 couplings;
(5) grammatical relation of Language data attribute 605 expression verb notions 601;
(6) time attribute of Time data attribute 606 expression verb notions 601, i.e. the time of origin of verb representative action;
(7) environment attribute of Environment data attribute 607 expression verb notions 601, i.e. the place that action takes place is feature extremely;
(8) hasPartOf object properties 608 expression eventClass class 602 is made up of the actConcept class, and the domain of hasPartOf object properties is eventClass classes, and range is the actConcept class;
(9) be set membership between hasSubClassOf object properties 609 expression actConcept classes and the actConcept class, the domain of hasSubClassOf object properties and range are the actConcept classes.
Claims (5)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2010102908605A CN101957812A (en) | 2010-09-21 | 2010-09-21 | Verb semantic information extracting method based on event ontology |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2010102908605A CN101957812A (en) | 2010-09-21 | 2010-09-21 | Verb semantic information extracting method based on event ontology |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN101957812A true CN101957812A (en) | 2011-01-26 |
Family
ID=43485145
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2010102908605A Pending CN101957812A (en) | 2010-09-21 | 2010-09-21 | Verb semantic information extracting method based on event ontology |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101957812A (en) |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102681982A (en) * | 2012-03-15 | 2012-09-19 | 上海云叟网络科技有限公司 | Method for automatically recognizing semanteme of natural language sentences understood by computer |
| CN103268311A (en) * | 2012-11-07 | 2013-08-28 | 上海大学 | Chinese Sentence Analysis Method Based on Event Structure |
| CN103365427A (en) * | 2012-03-31 | 2013-10-23 | 联想(北京)有限公司 | A method and electronic device for adjusting input content |
| CN106445990A (en) * | 2016-06-25 | 2017-02-22 | 上海大学 | Event ontology construction method |
| CN109885662A (en) * | 2019-03-01 | 2019-06-14 | 陈包容 | A kind of method of judgement event tense |
| CN110362656A (en) * | 2019-06-03 | 2019-10-22 | 广东幽澜机器人科技有限公司 | A kind of semantic feature extracting method and device |
| CN110612525A (en) * | 2017-05-10 | 2019-12-24 | 甲骨文国际公司 | Enable rhetorical analysis by using communication discourse trees |
| CN111985227A (en) * | 2020-07-20 | 2020-11-24 | 东南大学 | A method of automatic judgment of tense and voice in English sentences |
| US11960844B2 (en) | 2017-05-10 | 2024-04-16 | Oracle International Corporation | Discourse parsing using semantic and syntactic relations |
| US12001804B2 (en) | 2017-05-10 | 2024-06-04 | Oracle International Corporation | Using communicative discourse trees to detect distributed incompetence |
| US12141535B2 (en) | 2017-05-10 | 2024-11-12 | Oracle International Corporation | Techniques for maintaining rhetorical flow |
-
2010
- 2010-09-21 CN CN2010102908605A patent/CN101957812A/en active Pending
Cited By (15)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102681982A (en) * | 2012-03-15 | 2012-09-19 | 上海云叟网络科技有限公司 | Method for automatically recognizing semanteme of natural language sentences understood by computer |
| US9613311B2 (en) | 2012-03-31 | 2017-04-04 | Beijing Lenovo Software Ltd. | Receiving voice/speech, replacing elements including characters, and determining additional elements by pronouncing a first element |
| CN103365427A (en) * | 2012-03-31 | 2013-10-23 | 联想(北京)有限公司 | A method and electronic device for adjusting input content |
| CN103268311A (en) * | 2012-11-07 | 2013-08-28 | 上海大学 | Chinese Sentence Analysis Method Based on Event Structure |
| CN106445990B (en) * | 2016-06-25 | 2019-10-11 | 上海大学 | Event Ontology Construction Method |
| CN106445990A (en) * | 2016-06-25 | 2017-02-22 | 上海大学 | Event ontology construction method |
| CN110612525A (en) * | 2017-05-10 | 2019-12-24 | 甲骨文国际公司 | Enable rhetorical analysis by using communication discourse trees |
| CN110612525B (en) * | 2017-05-10 | 2024-03-19 | 甲骨文国际公司 | Enable rhetorical analysis through the use of communication discourse trees |
| US11960844B2 (en) | 2017-05-10 | 2024-04-16 | Oracle International Corporation | Discourse parsing using semantic and syntactic relations |
| US12001804B2 (en) | 2017-05-10 | 2024-06-04 | Oracle International Corporation | Using communicative discourse trees to detect distributed incompetence |
| US12141535B2 (en) | 2017-05-10 | 2024-11-12 | Oracle International Corporation | Techniques for maintaining rhetorical flow |
| CN109885662A (en) * | 2019-03-01 | 2019-06-14 | 陈包容 | A kind of method of judgement event tense |
| CN109885662B (en) * | 2019-03-01 | 2023-04-07 | 陈包容 | Method for judging event tense |
| CN110362656A (en) * | 2019-06-03 | 2019-10-22 | 广东幽澜机器人科技有限公司 | A kind of semantic feature extracting method and device |
| CN111985227A (en) * | 2020-07-20 | 2020-11-24 | 东南大学 | A method of automatic judgment of tense and voice in English sentences |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101957812A (en) | Verb semantic information extracting method based on event ontology | |
| Huang et al. | Multilingual twitter corpus and baselines for evaluating demographic bias in hate speech recognition | |
| Jha et al. | NLP-driven citation analysis for scientometrics | |
| Rajagopal et al. | A graph-based approach to commonsense concept extraction and semantic similarity detection | |
| Maynard et al. | Automatic detection of political opinions in tweets | |
| CN104933027B (en) | A kind of open Chinese entity relation extraction method of utilization dependency analysis | |
| CN110941953B (en) | Automatic identification method and system for network false comments considering interpretability | |
| CN106484664A (en) | Similarity calculating method between a kind of short text | |
| CN104268200A (en) | Unsupervised named entity semantic disambiguation method based on deep learning | |
| CN107832229A (en) | A kind of system testing case automatic generating method based on NLP | |
| El-Shishtawy et al. | An accurate arabic root-based lemmatizer for information retrieval purposes | |
| Lo et al. | An unsupervised multilingual approach for online social media topic identification | |
| CN107145514B (en) | Chinese sentence pattern classification method based on decision tree and SVM mixed model | |
| CN102508830A (en) | Method and system for extracting social network from news document | |
| Singh et al. | Writing Style Change Detection on Multi-Author Documents. | |
| CN116796751A (en) | A sarcasm detection method based on emotion-topic-sarcasm hybrid model | |
| CN107992550A (en) | A kind of network comment analysis method and system | |
| Parameswarappa et al. | Kannada word sense disambiguation using decision list | |
| Ryan et al. | A Comparative text similarity analysis of the works of Bernard Mandeville | |
| CN106503256A (en) | A kind of hot information method for digging based on social networkies document | |
| Xie et al. | Extracting lexical features from dialects via interpretable dialect classifiers | |
| Dung | Natural language understanding | |
| Tavan et al. | Identifying Ironic Content Spreaders on Twitter using Psychometrics, Contextual and Ironic Features with Gradient Boosting Classifier. | |
| Lamirel et al. | Federating clustering and cluster labelling capabilities with a single approach based on feature maximization: French verb classes identification with igngf neural clustering | |
| Patrick et al. | Developing SNOMED CT subsets from clinical notes for intensive care service |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C02 | Deemed withdrawal of patent application after publication (patent law 2001) | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Open date: 20110126 |