In the traditional Web, users are considered as information consumers. In social Web, users play a much more active role since they are now not only information consumers but also data providers. Users like online posting reviews which has become an increasingly popular way to express opinions and sentiments toward the products bought or services received. Analyzing these reviews can be helpful for collecting opinions of people about products, social events and problems and would produce useful actionable knowledge that could be of economic values to vendors and other interested parties. Thus, due to the huge number of reviews and their unstructured nature, efficient computational methods are needed for mining and summarizing these reviews, because regular analysis of reviews does not indicate user likes and dislikes. In a review, user typically writes about both the positive and negative aspects of the object, although the general sentiment toward that object may be positive or negative. That?s why sentiment analysis together with opinion mining try to extract and study of user?s opinions, sentiments and subjectivity of text. However, this analysis must come with careful consideration of user?s anonymity and the privacy of their sensitive data as privacy is today an important concern for both users and enterprises. In this research, automatic analysis of opinions (opinion mining) is performed to obtain such detailed aspects based on ontology. Opinion mining identify the features in the opinion and classify the sentiments of the opinion for each of these features. Opinion mining is a difficult task, owing to both the high semantic variability of the opinions expressed, and the diversity of the characteristics and sub-characteristics that describe the products and the multitude of opinion words used to depict them. In the proposed approach, the opinion polarity and polarity strength are measured using fuzzy set. As the fuzzy set theory is quite effective in processing natural languages, to measure the vagueness, it will also be effective in analyzing review articles, which are generally in natural languages. Additionally, the proposed system takes privacy into consideration by anonymizing data before final publishing. Methods of generalization and micro-aggregation are utilized for anonymizing quasi-identifiers to maintain the balance between data utility and user privacy.
展开▼
