Dissertation > Excellent graduate degree dissertation topics show

The similarity cut words and semantics of Chinese - based research and application

Author: PeiYunLiang
Tutor: GuXiaoFeng
School: University of Electronic Science and Technology
Course: Software Engineering
Keywords: <HowNet> semantic word similarity semantic string sentence similarity
CLC: TP391.1
Type: Master's thesis
Year: 2011
Downloads: 50
Quote: 0
Read: Download Dissertation

Abstract


In the Natural Language Processing (NLP), the text Similarity which is widely studied and discussed has important research value in many areas. The research of Chinese similarity uses linguistics, statistics, sociology and computer science to achieve computing the similarity of the various structure types of Chinese text. Since there is no natural division character in the Chinese text, therefore the research of Chinese similarity is premised on Chinese word segmentation. After years of exploration and research, there have some good effect automatic word segmentation systems in the Chinese word segmentation field. For example, the ICTCLAS system used in this dissertation is a Chinese word segmentation system with a good performance. Algorithms of similarity generally have the following categories: algorithm based on statistical models, algorithm based on rules and regulations and algorithm based on ontologies. All above of the algorithms have their own adervantages and disadervantages, however, the first two categories algorithms dosen’t consider the key semantic similarity in the comptutation of similarity, so the effect of them are not better than the last categories algorithms. Because of this point, in this dissertation we use the algorithm based ontology to research the Chinese word and sentence similarity.Based on the study and summarize previous works, the main work of this dissertation can be summarized as follows:1. In the dissertation firstly study the <HowNet> and similarity of semantic, for the problems of algorithms of the semantic similarity based on <HowNet>, a new algorithm of the semantic similarity is proposed with considering the main kinds of properties of semantic. Based on researching of <HowNet> and current algorithms of the words similarity, a new algorithm is proposed. Through the study of the co-occurrence of words, we believe that this relationship is also a reflection of the words’similarity, so the co-occurrence of words is introduced into the computation of the words similarity to amend the formular which is proposed in the dissertation. 2. Since <HowNet> dosen’t compute the similarity of unrecognized words, we use the word segmentation system and maximum metching algorithm to translate the unrecognized words to recognized words which are in <HowNet>, then the algorithm of unrecognized words similarity is proposed based <HowNet> and word segmentation.3. Based on the point of view that there having more then two words with high similarity in the words sequence,it would have high contribution for the similarity of sentence, the concept of the semantic string is proposed. We consider the weight of the semantic string to propose the new algorithm of sentence similarity.4. Based on algorithms of word and sentence similarity and using algorithm of VSM to compute the text similarity, we design and implement a full-text reterival system.

Related Dissertations

  1. Study on Noun’s Assertion in Zuo Zhuang,H146
  2. Research of Multiple Emails Automatic Summarization,TP391.1
  3. The Discovery of User Concept Region Based on Multiple Instance Learning,TP391.41
  4. Tourism Comments on the Internet’s Semantic Analysis and Usefulness Research,TP391.1
  5. A Study of Household Handles Design from the Perspective of Kansei Engineering,TS664.01
  6. An Approach for Business Process Oriented Service Dynamic Composition,TP393.09
  7. Semantic Retrieval Research Based on Ontology,TP391.3
  8. On George. Miller’s Cognitive Theory of Meaning,B842.1
  9. Ontology -based Semantic Web service matching and composition method,TP393.09
  10. WordNet and the \,G254
  11. The Frequency Characteristics of Mind Wandering: Questionnaire Development,B841
  12. Research of Intelligent Search Engine Based on Semantic Web,TP391.3
  13. The Study of the Range Adverbs in Mahasangha-vinaya,H141
  14. A Study of Category Theories in College English Vocabulary Teaching and Learning,H319
  15. Research on Topic Clustering Model of Socail Tagging Based on Bayesian Theory,C93
  16. \,H146
  17. Russian simple sentence semantics main function,H35
  18. English verbs hate comparative study psychology class,H314
  19. Semantic Structure of VCO Sentence in Modern Chinese,H146
  20. Synthetic Study of "Shanglai",H146
  21. An Analysis on the Chinese Four-character Construction "X Qian X Hou",H146

CLC: > Industrial Technology > Automation technology,computer technology > Computing technology,computer technology > Computer applications > Information processing (information processing) > Text Processing
© 2012 www.DissertationTopic.Net  Mobile