Abstract
When people want to purchase products or services, they usually acquire product or service documents to acquire opinions from costumers. People usually look for costumers with similar profiles and acquire their opinions. However, it is difficult for people to analyze various opinions and evaluate the values of distinct opinions. Moreover, people must spend much time and efforts to get representative opinions. On the basis of costumer’s opinions, our research develop a model for analysis and visualization of user profiles and opinions. Under the model, the free-form opinions can be converted into structured opinion components and the relationship between user profiles and opinions can be visually depicted. First of all, our research collects free-form opinions from internet or bulletin board system. Secondly, our research generalizes components of free-form opinions and analyzes the characteristics of these components. Then, the model automatically converts free-form opinions into formatted matrix and calculates of values corresponding to opinion tendency, user profile, and importance. Finally, the model visualizes the relationship between user profiles and opinions. By utilizing the model, this research develops a platform for opinion analysis and user profile determination. Furthermore, a real-word case, Moble01、ePrice、Yahoo!, are used to evaluate the feasibility and performance of the proposed model and platform. As a result, visualization of the relationship between user profiles and opinions can improve reading efficiency and effectiveness of users. On the whole, the proposed methodology and platform that automatically converts free-form opinions into structured opinion components and visualize the relationship between user profiles and opinions, then people can make quickly decision about the products and services based on the appropriate opinions of others.