Extractive Text Summarization for Social News using Hybrid Techniques in Opinion Mining

Presently almost all enterprises are oriented into building text data in abundance savoring the benefits of big data concept but the reality is that it’s not practically possible to go through all this data/documents for decision making because of the time constraint. Here in exists intense need of an approach as an alternative for the actual content which can summarize the complete textual content. By adopting these summarizing approaches, the accuracy in data retrieval of summarized content via search queries can be enhanced compared to performing search over the broad range of original textual content. There are many text summarization techniques formulated having their own pros and cons. The present work focuses on a comprehensive news review of extractive text summarization process methods and also taking into account, data appended dynamically. The existing work recommends a technique of hybrid text summarization that’s a blend of CRF (conditional random fields) and LSA (Latent Semantic Analysis) which being highly adhesive with low redundant summary and coherent and in-depth information. The above hybrid techniques is being extracted in five types that being: Positive and negative, statement, questions, suggestions and comments. The technique of LSA extracts hidden semantic structures within words/sentences that being commonly utilized in the process of summarization. The statistical modeling technique of CRF adopts ML (machine leaning) for offering structured detection and providing multiple options for evaluation of opinion summarization thereby identifying the most appropriate algorithm for news text summarizations considering the heavy volume of datasets.