4046

3824

كارشناسي ارشد

معماري كامپيوتر

اصفهان: دانشگاه صنعتي اصفهان، دانشكده برق و كامپيوتر

1386

ده، [121]، [II] ص.: مصور، نمودار

ص.ع. به: فارسي و انگليسي

محمد حسين سرائي

مازيار يالهنگ

18/04/87

علي محمد دوست حسيني، فريد شيخ السلام

مهندسي برق و كامپيوتر

ID3824

به فارسي و انگليسي: قابل رؤيت در نسخه ديجيتال

AbstractIn recent years vast amount of textual information is collected and stored in variousdatabases around the world including the Internet as the largest database of all Text hasturned into an important and critical resource for discovering information From businessto medical and security worlds for catching up with competitors to treat diseases and toprevent threats against national security necessity to extract information from the text or toestablish textual associations is undeniable This burgeoning growth of published textmeans that even the most avid reader cannot hope to keep up with all the reading in a field and nuggets of insight or new knowledge are at risk of languishing undiscovered in theliterature Text mining offers a solution to this problem by replacing or supplementing thehuman reader with automatic systems undeterred by the text explosion It involvesanalyzing a large collection of documents to discover previously unknown information The information might be relationships or patterns that are buried in the documentcollection and which would otherwise be extremely difficult if not impossible to discover Text clustering is one of the most important areas in text mining which includes textpreprocessing dimension reduction by selecting some terms features and finallyclustering using selected terms Feature selection is appeared to be the most important stepin the process Conventional unsupervised feature selection methods define a measure ofdiscriminating power of terms to select proper terms from corpus Evaluation of terms ingroups has not been investigated in reported works In thesis a new and robust unsupervised feature selection approach is proposed thatevaluates terms in groups Considering terms in group is to find terms that can group withother terms to be used for clustering based only on low power of discriminating Inaddition a new Modified Term Variance measuring method is proposed for evaluatinggroup of terms Furthermore a genetic based algorithm is designed and implemented forcomputing new measure and finding final feature vector use by clustering task In order toevaluate and justify our approach the proposed method and also conventional termvariance method are implemented and tested using corpus collection Reuters 21578 Results of comparing these two methods are very promising and show that our methodproduces better average accuracy and F1 measure than conventional term variance method

محمد حسين سرائي

مازيار يالهنگ

علي محمد دوست حسيني، فريد شيخ السلام