تطوير أنموذج تقييم السيارات على أساس المنطقة باستخدام التنقيب القائم على الرأي

القضاة, معتز نعيم (2019) تطوير أنموذج تقييم السيارات على أساس المنطقة باستخدام التنقيب القائم على الرأي. Masters thesis, جامعة الشرق الاوسط.

Text 5d47d2814f7c8_1.pdf Download (2MB)

Abstract

With the increasing expansion of social-network and e-commerce sites, many people have started to rely on the opinions in these Websites as a reference when purchasing a product or service, also, helping them to make purchasing decisions. Because of the many reviews available in the Social-Networks on various topics it is difficult to read all of comments and opinions. So, this study aimed to reviewing many previous models in the field of SentimentAnalysis and Opinion-Based on the people reviews in social-network sites and proposing a developed model that takes Region-Based as important variable of the search process Due to different opinions about the same product in different regions or countries, and the researcher found a gap in the previous models because it did not cover this important aspect The study adopted the method of Systematic Analysis, the researcher collected the sample of the study from Twitter by using TwitterR and web crawling. The sample of the study, which was collected from two countries, United Kingdom and United State of America, and each sample was 125 tweets containing the reviews about their opinion of cars, specifically “Toyota Rav4”. Then we make data Pre-Processing to clean the data. The next step is Sentiment-Analysis for each review and rate it as positive, negative or neutral and give a rate from -5 to +5 for each review. After this phase, the rating average of all reviews for each country has been taken. The results show that there are differences in the opinions for the same product based on region. Where the results showed that the average users’ ratings in UK is 2.02 and in USA is 0.44. We can measure the validity of the study tools by considering all the steps followed in the model have been applied in previous studies and researches. ن The study recommends expanding the data collection to gather largest number of users reviews in social-network and developing special software’s for Sentiment-Analysis and opinion-based Mining to support Arabic language, and eventually applying the model in different sectors of products and services. Keywords: Sentiment-Analysis, Opinion-Based Mining, social-network, users reviews rating, car rating.

Actions (login required)

View Item