Paper Title
Hybrid Text Features Extraction System using Multilayer Bidirectional LSTM and PSO Algorithm

Presently, with data as a trend, we have been bombarded with a huge amount and variety of data and with this tremendous increase in the data and availability of information from all over the place, it becomes humdrum to deal with the appositeness of the information received in the form of a variety of documents, articles or drafts. There can be a constant fear of missing out on some of the most important information if proper scrutiny is absent. Making this huge data into its smaller form can prove as a great help in understanding the data, its pattern, and relevance at the same time. Doing this task manually can be troublesome, time consuming and even inefficient in some cases. Thus, we need a technique to make these lengthy texts and documents short and precise. Text summarization is one such technique that helps in selecting useful information and simultaneously maintains the quality and accuracy of the data keeping its real context. When we talk about text summarization, it can be achieved using the extractive approach and abstractive approach. In extractive based text summarization, we usually use weights of the important and relevant parts of the sentence and use the total weights as resultant to obtain the concise version of our text or document. Whereas in abstractive based text summarization, we used some advanced deep learning techniques like Multilayer Bidirectional lstm to obtain the summary. The main agenda of this paper is to focus on the development of a hybrid method to extract the features using tfidf from a given summary and obtaining optimised tokens out of it using Particle Swarm Optimization algorithm. Keywords - Extractive, Abstractive, Term Frequency Inverse Document Frequency Algorithm, Multilayer Bidirectional, Lstm, Particle Swarm Optimization Algorithm.