Dissertation > Excellent graduate degree dissertation topics show
A Study of Algorithms for Text Categorization Based on Reducing Class and Fuzzy Theory
Author: ZengHongBo
Tutor: YangTianQi
School: Jinan University
Course: Applied Computer Technology
Keywords: Category cutting Fuzzy Theory Text Classification KNN Class skew
CLC: TP391.1
Type: Master's thesis
Year: 2011
Downloads: 37
Quote: 0
Read: Download Dissertation
Abstract
|
Text classification has become a key technology for processing large amounts of text information , and has become an important research directions in the field of data mining . KNN text classification algorithm is one of the many automatic text classification techniques in performance relative prominence , at the same time , it also has its own inadequacies , the first classification slow KNN classification algorithm to All calculations are deferred to the classification , and every must calculate the test text classification similarity of text with all the training , classification slow ; second , the class skew the processing of the data set is not ideal , is much larger than the number of another class training documents some categories , the number of text , the classification results would tend to text a large number of categories . For these two shortcomings , this paper proposes the corresponding solutions . Proposed a method based on class cutting , improve the speed of KNN classification , the program text classification is divided into multi-step , every step through the rapid classification cropped some categories until the last remaining category , classified as the final results. Second, the fuzzy theory is applied to the KNN algorithm for training text set and set of categories to establish a membership degree matrix , then this matrix to calculate the class attribute of the test text . And a balanced membership matrix factor , thereby reducing the negative impact of the difference in the number of category text classification results . The experiments show that the proposed scheme achieved the desired results , the former can greatly reduce the classification time does not affect the classification results , ; latter KNN classification performance can be significantly improved , especially in dealing with the class - skewed data set The performance improvement is more significant .
|
Related Dissertations
- Research on Dam-Break Probability Analysis Method for Dangerous Reservoir,TV697
- Research on Text Classification Based on Biomimetic Pattern Recongnition,TP391.1
- Tourism Comments on the Internet’s Semantic Analysis and Usefulness Research,TP391.1
- Based on Data Distribution Characteristics of Text Classification,TP391.1
- Research on Improved K Neighbor Support Vector Machine Algorithm Faced Text Classification,TP391.1
- Research of Image Enhancing of Vein Based on Fuzzy Theory,TP391.41
- Research for Event Extraction Method in Specific Domain Based on Tree Conditional Random Field,TP391.1
- Online Education News Text Categorization System Design and Implementation,TP391.1
- One kind of empirical data on the workload of a software bug fixes Prediction Model,TP311.53
- Ontology-based medicine named entity recognition technology research,TP391.1
- Research on cross-language text categorization,TP391.1
- Based on Fuzzy Clustering of Web users,TP311.13
- Classification model based monitoring of e-commerce Prohibited Research and Implementation,TP393.09
- Based on semantic analysis of text mining research,TP391.1
- Analysis on Stability and Construction Mechanical Behavior of Karst Tunnel Surrounding Rock,U451.2
- Planning and Schedling of Production Transferring Projects of International Company,F276.7;F273
- Based on the associated technology Chinese Text Classification,TP391.1
- Automatic distribution of software defects,TP311.52
- Research of Web Text Classification Based on Decision Tree Classification Algorithm,TP391.1
- Study on Chinese Text Classification Combined with Ontology,TP391.1
- Spam messages based on ensemble learning multi-level classification technology research,TN929.53
CLC: > Industrial Technology > Automation technology,computer technology > Computing technology,computer technology > Computer applications > Information processing (information processing) > Text Processing
© 2012 www.DissertationTopic.Net Mobile
|