Sentiment Analysis of Users for Buying on Cell Phone in Digikala

Abstract

Nowadays, people use others' opinions on social networks for decision-making to purchase online products and services. Likewise, the companies which offer the products employ sentiment analysis of opinions of users and customers to adopt informed decisions and offer new products. Considering the high volume of the contextual data, conversion, and analysis of such data is a major challenge in e-commerce. Sentiment analysis is a modern approach in the extraction of opinions. The obtained information from sentiment analysis can have a considerable impact on the efficient selection of customers. In the present study, a model has been proposed for sentiment analysis of users' opinions for buying a cell phone in Digikala. This study is applicable to the objective aspect. The data includes users' opinions in Digikala. The statistical sample consists of opinions of cell phone users in Digikala. Supervised learning, as well as Python package, were utilized for analysis and implementation. A model has been proposed for sentiment analysis of users' opinions. The results demonstrate that this model can classify users' opinions with an accuracy equal to 0.892. Similarly, the results reveal that users' opinions about ease of use, possibilities, and capabilities of the cell phone are positive and about purchase value to price, innovation, design and appearance, and quality of cell phones are negative. The proposed model can be implemented in e-commerce websites like Digikala and its output can be observed by users systematically. Finally, it can be led to inform decision-making for buyers and companies which offer products.

Keywords

• sentiment analysis
• social networks
• supervised learning