Sentiment analysis algorithm for Arabic reviews on the movies domain

العناوين الأخرى

خوارزمية لتحليل المشاعر للتعليقات العربية على نطاق الأفلام

مقدم أطروحة جامعية

al-Mansuri, Khalid Walid Nassar

مشرف أطروحة جامعية

al-Shalabi, Riyad

أعضاء اللجنة

al-Shaykh, Asim
al-Mashaykhi, Akram

الجامعة

جامعة عمان العربية

الكلية

كلية العلوم الحاسوبية و المعلوماتية

القسم الأكاديمي

قسم علم الحاسوب

دولة الجامعة

الأردن

الدرجة العلمية

ماجستير

تاريخ الدرجة العلمية

2017

الملخص الإنجليزي

Sentiment analysis has become popular in recent years and has been used in many life aspects where the people reviews are very important to know their reaction on any product.

Regarding movies, these impressions are important for people to choose the best film to watch.

Movie reviews are a very important tool for measuring the performance of the movie, most of the sentiment analysis research in those reviews are in English language, but our goal is how to analyze the feelings and reviews using Arabic language.

In this thesis, an Arabic lexicon for movies was created and the lexicon is used in applying movies sentiment analysis.

Movie Rating for Arabic Reviews System ( MRARS ) was built for sentiment analysis for Arabic movies reviews using Oracle Language.

The lexicon contains all the feelings words about movies which are positive, negative and neutral sentiment.

We collected about 1250 words, the number of positive words is 655, the number of negative words is 441 and the number of neutral words is 154.

The proposed system showed excellent performance varied between 94% - 100% and the accuracy can be raised after adding new words to the lexicon.

التخصصات الرئيسية

تكنولوجيا المعلومات وعلم الحاسوب

الموضوعات

• العواطف
• الخوارزميات
• اللسانيات الحاسوبية

عدد الصفحات

63

قائمة المحتويات

• APA
• MLA
• AMA

لغة النص

الإنجليزية

نوع البيانات

رسائل جامعية

رقم السجل

BIM-764291