Dissertation > Excellent graduate degree dissertation topics show

Cluster Analysis of Microblog Topic's Comments

Author: ZhangChao
Tutor: WangXiaoLong
School: Harbin Institute of Technology
Course: Computer technology
Keywords: subtrees analysis evaluation pattern mining semantic dictionary similarity calculation
CLC: TP393.092
Type: Master's thesis
Year: 2012
Downloads: 27
Quote: 0
Read: Download Dissertation

Abstract


Current micro-blogging push mode has many shortcomings, such as toomuch spam, inconvenient micro-blog reading resulted from too many pages,invalid opinion reading. The emphasis of this thesis focuses on how to overcomethe problem of current micro-blogging push mode and finding a more humane,convenient way to provide micro-blog opinion for users.The thesis designed and implemented a micro-blog public opinion analysissystem. Firstly, the result of dependency parsing treatment was analyzed usingthe subtrees analysis method. Secondly, the thesis used the method of evaluationpattern mining to mine results of the subtrees analysis and got key words ofevaluations, then replaced original evaluations with key words of correspondingevaluations. Thirdly, all topics and evaluations are clustered using the improvedsimilarity calculation method, and each clustering result represented one themeor one evaluation.The contribution of this paper is as follows:(1) The subtrees analysis method is implemented. The method divides onesentence into several sub-sentences, splits each sub-sentence into some partsaccording to its relatively independent meanings. Each part represents itsspecific syntactic role.(2) One mining method of evaluation pattern for views or opinions isproposed. The method uses Apriori algorithm to mine the result of the subtreesanalysis and gets some patterns, then defines some types of evaluation patternsaccording to observation of patterns and knowledge of Chinese characteristicsand gets some relatively fixed evaluation patterns using statistical methods,finally applies the above knowledge of evaluation patterns to mine the key wordsof evaluations and gets key words of evaluations.(3) One improved similarity calculation method is proposed. The methodachieves the shift from traditionally computing a single word’s maximumsimilarity to calculating the whole sentence’s maximum similarity. Someexperiments on four kinds of sentence analysis methods show the effectivenessof the improved method.

Related Dissertations

  1. Research of Multiple Emails Automatic Summarization,TP391.1
  2. Study and Implement on Spherical Harmonc Based 3D Models Retrieval System,TP391.41
  3. Study on Data Mining in Water Dispatching Decision Support System,TV697.11
  4. A Conceptual Query Based Multi-Document Summarization in Biomedical Domain,TP391.1
  5. The Research of an Electronic Commerce Recommendation System Based on Semantic Dictionary,F713.36
  6. Reasoning knowledge retrieval method based on ontology and SWRL,TP391.3
  7. Based on the writing style characterized plagiarism checking Optimization Method,TP391.1
  8. Design and Implementation of Jee-based Scientific Project Management System,TP311.52
  9. Body in the integration of medical information and research,TP391.1
  10. Chinese text orientation Classification System,TP391.1
  11. Research and Implementation of Chinese Question-Answering System Based on Multi-Agent & Meta Search Engine,TP391.1
  12. Design and Implementation of Fingerprint Spectrum Differenece Analysis System of the Complex Multi-component System,O657
  13. Studies of Query Expansion Based on Semantic Dictionary and Local Analysis,TP391.3
  14. Content Analysis Based Patent Mining Research,TP391.1
  15. Research and Application on Semantic Retrieval for Digital Library Based on Ontology,G250.76
  16. Research on the Content-Based Piano Music Retrieval,TP391.3
  17. Research of Image Retrieval Based on Dual-tree Complex Wavelet Transform and Information Entropy,TP391.41
  18. A Concept Similarity Computation Based on Multi-Property Ontology,TP391.1
  19. Research and Implementation on Modeling Temporal Relation of Event,TP391.1
  20. Topic Tracking of Accidental News Based on SVM,TP391.1
  21. Research on the Algorithm for Chinese Duplicated Web Pages Detection,TP393.092

CLC: > Industrial Technology > Automation technology,computer technology > Computing technology,computer technology > Computer applications > Computer network > General issues > The application of computer network > Web browser
© 2012 www.DissertationTopic.Net  Mobile