Dissertation > Excellent graduate degree dissertation topics show

Technology Research on Sentiment Analysis for Chinese Web Reviews

Author: ZhouCheng
Tutor: XiaoWeiDong
School: National University of Defense Science and Technology
Course: Management Science and Engineering
Keywords: Chinese Web Reviews Sentiment Analysis Text Classification Emotional Dictionary Emotional tendencies
CLC: TP391.1
Type: Master's thesis
Year: 2011
Downloads: 261
Quote: 2
Read: Download Dissertation

Abstract


With the rapid development of network technology, the network has become an important source of more and more people access to information, but also become a platform for people to express their views. Excavation and analysis of the rapid growth of online text resources, in particular the initiative to publish reviews, identify their emotional tendencies and evolution, you can better understand the behavior of the user, the analysis of hot public opinion, but also for the government, enterprises and other institutions provide an important basis for decision-making. This paper first introduces the research background and sentiment analysis applications prospects; then as the research object to Chinese Web Comments, introduced its concept, characteristics; sentiment analysis process Next, according to Web reviews, comments from Web access and pretreatment web comments sentiment analysis method, two aspects of the in-depth study. Web Comments emotion analysis based text classification techniques and sentiment analysis method based on emotion dictionary. The value of sentiment analysis is to come to the conclusion of the summary analysis from the reviews of a particular topic, which involves first reviews data network. Reviews of the same subject are usually concentrated in certain sites, the pages of the same site is highly structured. For this feature, we design a web-based messaging middleware, real-time processing technology to parallel download and pretreatment page, review data for sentiment analysis. Then, we use two kinds of the emotional analysis methods based on different ideas: (1) based text classification techniques: First, in the traditional feature selection method based on the proposed selection algorithm, based on the joint correlation and redundancy features to delete redundant I characteristics, reserved in favor of classification features, thereby enhancing the the text emotion classification effect; final text classification using support vector machine method of emotional polarity classification. (2) based on emotion dictionary technology: HowNet establish an emotional dictionary, and calculate the emotional tendencies of the Chinese words and calculate the the the emotional tendencies value phrases in the text, and then based on the phrase structure further, finally obtained by summing the entire comment emotional tendencies value . Finally, manually labeled data sets get to comment publicly on the network data sets and subject to experimental test data, comparative analysis of the proposed two sentiment analysis methods, experimental results show that: two sentiment analysis method presented in this paper are effective and emotional dictionary-based approach in the performance slightly better than the text-based classification approach.

Related Dissertations

  1. Research on Text Classification Based on Biomimetic Pattern Recongnition,TP391.1
  2. Tourism Comments on the Internet’s Semantic Analysis and Usefulness Research,TP391.1
  3. Based on Data Distribution Characteristics of Text Classification,TP391.1
  4. Research on Sentiment Orientation Analysis of Blog Article Based on Blog Search,TP391.1
  5. Research on Improved K Neighbor Support Vector Machine Algorithm Faced Text Classification,TP391.1
  6. Research for Event Extraction Method in Specific Domain Based on Tree Conditional Random Field,TP391.1
  7. Online Education News Text Categorization System Design and Implementation,TP391.1
  8. One kind of empirical data on the workload of a software bug fixes Prediction Model,TP311.53
  9. Data mining technology in the Expo traffic with social media Prediction Research and Application,TP311.13
  10. Research on cross-language text categorization,TP391.1
  11. Classification model based monitoring of e-commerce Prohibited Research and Implementation,TP393.09
  12. Figure based on association rules and emotional tendencies sort of sentence,TP391.1
  13. Based on semantic analysis of text mining research,TP391.1
  14. Based on the associated technology Chinese Text Classification,TP391.1
  15. Automatic distribution of software defects,TP311.52
  16. Research of Web Text Classification Based on Decision Tree Classification Algorithm,TP391.1
  17. Study on Chinese Text Classification Combined with Ontology,TP391.1
  18. Spam messages based on ensemble learning multi-level classification technology research,TN929.53
  19. Spam Filtering System Study Based on Non-negative Matrix Factorization,TP393.098
  20. News Web Texts Classification Based on Contents,TP391.1
  21. Research and Implementation on Key Technology of Web Text Collection and Analysis,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