Particle swarm optimization model for Hindi text summarization

Text Summarization is one of the techniques that shorten the original text without vanishing its information as well as meaning. A lot of algorithms exist for text summarization. Two approaches namely Abstractive Text Summarization and Extractive Text Summarization are used for this purpose. In Abstractive text summarization, the entire document is regenerated using a few lines. Whereas in Extractive Text Summarization sentences are filtered based on some ranks assigned to them by a specific algorithm. A lot of work has already been done in languages like English, Chinese etc. In this paper, the authors propose the summarization of Hindi text using the Particle Swarm Optimization model. Initially, the text in the Hindi language is summarized using a ranking-based technique then PSO (Particle Swarm Optimization) is applied to have an optimized summary of the text. One of the ranking-based techniques i.e. TF-IDF is introduced. Implementation of the proposed Systems is initially discussed in five steps- preprocessing, feature extraction, ranks generation, post-processing and optimized summarization using PSO. At the end, results are shown in terms of an optimized summary of text in a specific language. This system can be implemented in any standard language, but Hindi is selected for practical implementation because very few research work is done in the Hindi Language.