Dissertation > Excellent graduate degree dissertation topics show

Research of Patent Document Analysis and Retrieval Based on Latent Semantic Analysis

Author: XuYuanHao
Tutor: SunShouQian；SunLingYun
School: Zhejiang University
Course: Applied Computer Technology
Keywords: Patent analysis text mining latent semantic analysis patent clustering patent retrieval
CLC: TP391.3
Type: Master's thesis
Year: 2010
Downloads: 268
Quote: 4
Read: Download Dissertation

Abstract


Patent documents contain important research results, the content of which is wide-ranging and technical details are described in detail. Patent documents are significant sources of the latest information and technology in the world. Effective analysis of the patent document can improve the competitiveness of the enterprise in market.Based on the analysis of existing patent analysis techniques in the world, this thesis studied how to use text mining technology to the Chinese patent document analysis process. Used Latent Semantic Analysis and SOM network to cluster patent documents, and developed the corresponding software platform.Currently there is no patent text corpus in China. This thesis described the procedures of automatically download patent full-text from website. Due to the special nature of patent documents, the result of using traditional Chinese word segmentation techniques on patent document is not good. This thesis designed a new algorithm to patent new word identification to improve the results.It is an effective means for patent analysis with text mining. Traditional clustering method can only apply in low-dimensional objects, but in the face dimension of high-dimensional objects such as text, clustering methods cannot get good results. This thesis use Latent Semantic Analysis to reduce the dimension of patent document and kept the original semantic space structure. Then patent documents were clustered using SOM algorithm. Experimental result showed that the clustering time of dimension reduction text is less than original text, and the clustering result was good.In conventional patent search web sites the scope of search is patent abstract and not the full text. Full-text patent search engine was developed based on Lucene. Patent full texts were indexed by inverted index structure to speed up the retrieval time. The system can rank the search result according to relevance between document and query term, which can effectively reduce the amount of patent documents for user to view and improve efficiency

Related Dissertations

  1. Applications of Bibliometrics and Text Mining in the Life Science,TP391.1
  2. Research on Sentiment Orientation Analysis of Blog Article Based on Blog Search,TP391.1
  3. Based on the model and features of the package ML_pLSA Target Recognition Algorithm,TP391.41
  4. Chinese Movies in the Vision of American Movies Critics,J905
  5. Text -oriented disciplines correlation analysis association rule mining technology research,TP311.13
  6. Semantic-based Information Retrieval System Technology Document Design and Implementation,TP391.3
  7. Research of WEB Mining Applying in the Monitoring System of the Network Policing Information,TP393.09
  8. Finding Web Services Based on Clustering Probabilistic Semantic Approach,TP393.09
  9. Research on Approaches of the Subjective Automated Assessment,TP391.1
  10. Research on Co-clustering and Application,TP311.13
  11. The Study of Topic-Oriented IT News with Search Enging and Web Page Analysing,TP393.092
  12. Research on Topic Clustering Model of Socail Tagging Based on Bayesian Theory,C93
  13. Study on Hierarchical Text Categorization of Patent Data Based on Fuzzy Logistic,TP311.13
  14. Some of the key technologies of Internet text data mining,TP311.13
  15. Study on University-Industry Cooperation Mode and Characteristics Based on Patent Data of "985 Project" Universities of the Yangtze-Delta,G649.2
  16. Text-based mining companies Distress,F426.82
  17. Patent application Strategy,F426.4;D923.42
  18. Based on patent cited the influence of enterprise technology innovation and strategic analysis,F224
  19. Research on Text Clustering Methods and Their Applications,TP391.1
  20. Biomedical Named Entities Recognition Based on Classifiers Ensemble,TP18

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