Dissertation > Excellent graduate degree dissertation topics show

Research and Application of Chinese words and sentences similarity algorithm

Author: LiuQingLei
Tutor: GuXiaoFeng
School: University of Electronic Science and Technology
Course: Computer Software and Theory
Keywords: HowNet Word similarity Sentence similarity Total Information Variance Information
CLC: TP391.1
Type: Master's thesis
Year: 2011
Downloads: 80
Quote: 0
Read: Download Dissertation

Abstract


The similarity of words and sentences in information retrieval, text categorization, question answering system, based on a wide range of applications in the field of machine translation instance. As the focus of the current discussion, the paper mainly from the perspective of semantic similarity calculation method based on HowNet word and sentence has conducted in-depth research. Calculated based on HowNet words (sentences) similarity usually sememe (words) between the optimal matching the final overall similarity by the similarity of each part as the basic unit of operation, through appropriate weighting calculation synthesized from such practices often result in repeated and unreasonable structure of some matching internal information. This thesis is precisely the problem, in-depth analysis of the HowNet architecture based on similarity calculation of words and sentences, and word sense disambiguation innovative and exploratory research. Mainly as follows: 1. Detailed study HowNet-based, meaning the original similarity. Such the sememe similarity computer system is often obtained by calculating the distance of the two sememe path in the upper and lower levels of relationship tree similarity between the meaning of the original, no or very few consider sememe depth in the hierarchy tree, considering the distance of the sememe the path well sememe depth of the hierarchy tree, From the sememe contains information on the number of nodes, a new meaning of the original similarity calculation method. The detailed study similarity calculation method commonly used words and phrases, and discussion focused on the similarity calculation based on the words and sentences of HowNet above. The article is from the perspective of information theory, sememe similarity algorithm based on the above statistics, which come two the righteousness original collection of Total and difference information between the two the righteousness original collection The similarity between the final word (sentence) the similarity of the original collection is the unit of calculation is based on justice. 3. Study the basic principles and specific method of word sense disambiguation algorithms currently used. HowNet provided with examples of well above sentence similarity algorithm based on a simple and efficient word sense disambiguation method. Specific analysis of automatic question answering system to build the technology and modular architecture, based on the model of FAQ answering system set for instance, reflects the importance of words and sentence similarity in specific applications.

Related Dissertations

  1. Research of Text Clustering on Food Complaint Documents Based on Ontology,TP391.1
  2. Tracking Events for Food Complaint Documents Based on Ontology,TP391.1
  3. Study on Chinese Text Classification Combined with Ontology,TP391.1
  4. Research on Web Content Filtering Based on Concepts of Collection,TP393.092
  5. Research of Single-document Summarization Based on Semantics,TP391.1
  6. Word Sense Disambiguation Technology Research Based on HowNet and Bayesian Model,TP391.1
  7. Research on Chinese Semantic Dependency Analysis,TP391.1
  8. Ellipsis Recovery Technology in Interactive Question Answering System,TP391.1
  9. Study on Chinese Semantic Orientation Analysis Based on HowNet,TP391.1
  10. The Research of Compliance Testing Technology of Traffic Terminology and Standards,TP391.1
  11. Research on Paper Similarity Based on Semantic Understanding,TP391.1
  12. The Study of Ontology-Based Query Expansion,TP391.1
  13. Research on Cross-lingual Word Similarity Computation,TP391.1
  14. Semantic Web - based personal digital site implementation,TP391.1
  15. Title Analysis and Application Based on Entity-Attribute Frame,TP391.1
  16. Research of Approaches to Hownet Based Chinese Word Sense Disambiguation,TP391.1
  17. In the name of the entity body application of information extraction,TP391.1
  18. Chinese word semantic similarity measure its cross-language information retrieval,TP391.1
  19. Research of Ontology-based Semantic Retrieval Model,TP391.3
  20. Research on Aligning Visual Properties and the Corresponging Describing Words of Images,TP391.41
  21. The Research and Implementation on Question Understanding and Similarity Computation of Chinese Question Answering System,TP391.1

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