الفهرس 
chapter one introductionOnly 14 pages are availabe for public view
 |  | from | 152 |  |  |
| | | from | 152 | | |
Abstract
The emergence of the Internet has changed the world drastically. As a result, everyday life has changed as well. Additionally, the development in the electronics industry helped the internet to dominate almost every aspect of life. Also, the telecommunication revolution has made devices cheap and now anyone in the world is able to have his own telecommunication device such as cell phone, computer or tablet with the least possible cost. Consequently, the world economy has changed and unless business takes benefit from this new technology, the business is going to fade away. Therefore, there is a need for business activities to be adapted in order to cope with this technology. One of the major business activities is marketing. Companies should develop marketing activities to cope with the digital world. Therefore, internet marketing emerged. Due to the limitation of the business budget, there is a need to precisely target the potential customers online otherwise, the budget will be wasted. With the proliferation of the information online, it is hard for people to find what they need, and hard for companies to target their potential customers. Therefore, there is a need for a system that helps people find what they need easily, and as a result, the companies can achieve their goals. This dissertation proposed a recommendation system model that helps users to find what they need in a quick time. In order to achieve that goal, three studies were conducted through the dissertation. The first study was conducted to extract the emotions and feelings from the text that users wrote online as reviews about the products. The output of this study was used in building the proposed model. The second study was to build a hybrid recommender system that takes into consideration the emotional fingerprint model which uses the user’s emotions to help him in making a decision. The third study was to add an upgrade to the proposed recommender in order to take into consideration the contextual factors such as the user’s mood, the geographical location, the time constraints, and the emotional factors as well. Experiments for the studies have been conducted and the results were measured using the standard evaluation metrics to find the performance of the proposed recommenders. Additionally, different studies were used as a baseline to compare the proposed work with them. The proposed approach outperformed the used baseline with a precision metric of 0.92.