Enhanced Discoverability of Content through Linked Data for Online Reviews using Classification and Ranking Techniques

Massive unstructured data are available and being posted in numerous blogs, forums, and online sites. This enormous amount of information on worldwide network platforms make them feasible and can be used as input source, in applications based on opinion mining and sentiment analysis. The aim of this paper is to analyze online reviews in unstructured form and discover content through linked data and deriving an opinion. Our proposed methodology comprises of phases such as Data pre-processing, content discovery and Opinion mining. Initially the unstructured data is extracted from the web document. This phase is used for formatting the data before sentiment analysis and mining. The second phase will be classified into two i.e., Feature extraction, content discovery and opinion extraction. The features like term frequency, Part of Speech are extracted from the words in the documents. After feature extraction, we extract useful information related to the item’s features and use it to rate them as positive, neutral, or negative. This final phase will be done by supervised learning algorithm decision tree classifier with the help of features extracted. In the final step ranking and classification will be done