![](/images/graphics-bg.png)
Feature based approach in Arabic opinion mining using ontology
Other Title(s)
التنقيب عن الآراء العربية باستخدام الانتولوجيا بالاعتماد على المستوى
Dissertant
Thesis advisor
University
Islamic University
Faculty
Faculty of Information Technology
Department
Department of Information Technology Systems
University Country
Palestine (Gaza Strip)
Degree
Master
Degree Date
2016
English Abstract
With the rapid increase in the volume of Arabic reviews that use applications such as online review sites, blogs, forums, social networking, and so forth, comes at an increasing demand for Arabic opinion mining techniques.
In Arabic language, researchs in this area is progressing at a very slow pace compared to that being carried out in English and other languages.
In this thesis, we highlight two problems for Arabic opinion mining technique: firstly, when analyzing review having different features with diverse opinion strengths.
It considers all features extracted from the reviews to be equally important in failing to determine the proper polarity of the review and makes the review’s sentiment classification less accurate.
Secondly, the opinion summary for each feature doesn’t consider the sub-features that represented it and makes the feature-based summary is incomplete.
This research presents a technique using ontology that work at feature level classification to classify Arabic user generated reviews by identifying the important features from the review based on level of these features on the ontology tree and to generate an opinion summary for each feature in the whole corpus by identifiying the opinion of its sub-feature terms in the ontology.
To evaluate our work, we use public datasets which are hotels and books datasets.
We use accuracy, recall, precision, f-measure metrics to evaluate the performance and compare the results with other supervised or unsupervised techniques.
Also, subjective evaluation is used in our method to demonstrate the effectiveness of feature and opinion extraction process and summarization.
We show that our method improves the performance compared with other opinion mining classification techniques, obtaining 78.83% f-measure in hotel domain and 79.18% in book domain.
Furthermore, subjective evaluation shows the effectiveness of our method by obtaining an average f-measure of 84.62% in hotel domain and 86.31% in book domain.
Main Subjects
Information Technology and Computer Science
No. of Pages
78
Table of Contents
Table of contents.
Abstract.
Abstract in Arabic.
Chapter One : Introduction.
Chapter Two : Theoretical background.
Chapter Three : Literature review.
Chapter Four : Research methodology.
Chapter Five : Experiments and results.
Chapter Six : Conclusions and future works.
References.
American Psychological Association (APA)
al-Asmar, Ahmad Muhammad I.. (2016). Feature based approach in Arabic opinion mining using ontology. (Master's theses Theses and Dissertations Master). Islamic University, Palestine (Gaza Strip)
https://search.emarefa.net/detail/BIM-707733
Modern Language Association (MLA)
al-Asmar, Ahmad Muhammad I.. Feature based approach in Arabic opinion mining using ontology. (Master's theses Theses and Dissertations Master). Islamic University. (2016).
https://search.emarefa.net/detail/BIM-707733
American Medical Association (AMA)
al-Asmar, Ahmad Muhammad I.. (2016). Feature based approach in Arabic opinion mining using ontology. (Master's theses Theses and Dissertations Master). Islamic University, Palestine (Gaza Strip)
https://search.emarefa.net/detail/BIM-707733
Language
English
Data Type
Arab Theses
Record ID
BIM-707733