الفهرس | Only 14 pages are availabe for public view |
Abstract Sentiment analysis or opinion mining is used to automate the detection of subjective information such as opinions, attitudes, emotions, and feelings. Sentiment analysis becomes an important source in people trust. As search engine land statistics and bright local survey, 92% of consumers trust online reviews as much as personal recommendations. Many researchers spend long time searching for useful papers. Online reviews on papers are the essential source to help them. There are several challenges that are obstacles in sentiment evaluation. The difficulty of understanding computer human sentences or linguistics. Thus, online sentiments can save the researcher’s time, it provides effort and paper cost. Analyzing scientific papers domain is hard. Evaluating sentiments with respect to several properties is hard. This domain requires scientific lexicons for parameters or features. In this thesis, we propose a new technique to analyze online reviews in the scientific research domain called: sentiment analysis of online papers (SAOOP). SAOOP aims to support researchers and save their time and effort by enabling them to report the total evaluation for the papers. SAOOP introduces a hybrid model and creates a new criteria for evaluating scientific papers. This hybrid model of an enhanced bag-of-words and Part-of-Speech models. SAOOP improves accuracy with solving several sentiment challenges. SAOOP consists of two evaluations for each research paper: Sentiment score and system score. Sentiment score is an evaluation of online sentiments. System score is a new criteria of evaluation topic parameter |