నైరూప్య

Development of an Enhanced Efficient Parallel Opinion Mining for Predicting the Performance of Various Products

K.NATHIYA, Dr.N.K.Sakthivel

with the help of Web 2.0, the centers around user participation, posting online reviews has become an increasingly popular way for people to share their views with other user’s opinions and sentiments toward products and services. It becomes a common practice for e-commerce websites to provide the facilities for people to communicate and publish their reviews between them. These online reviews present a wealth of information on the Services and Products, which will facilitate the improvement of their business. Hence a growing number of recent studies have been focused on the Opinion Mining. A few Opinion Mining based methods have been studied and analyzed. From our study, it is observed that a few opinion mining based TSCAN algorithm had not produced good results due to referring the users and customers opinions features with similar meaning as different. To overcome this issue, the Modified TSCAN Algorithm is proposed here. It is mainly focusing on the experts opinions which overcomes the existing system drawback in terms of referring genuine opinions, so that the readers could understand the content easily. By using this model, the more information can be extracted and associated through their temporal closeness, which will give comprehensible content. This model is involving vital role in the Opinion Mining because users can share their opinions about the products. From our implementation, it is observed that this scheme provides the best suitable solution for the user’s interests and demands. Thus our research work is proposed and implemented an efficient method for Opinion Mining called an Enhanced Efficient Parallel Opinion Mining (EEPOM) based TSCAN Algorithm. It is focusing more websites and it is extracting more information in parallel manner, so that we can get optimized fruitful result with the expert’s opinions. From our results, it is noted that it provides the best suitable solution for the user’s interests and demands and it i s improving the performance of existing technique in terms of Quality of Information, Prediction and obtaining of genuine opinions.