Multilingual Text Summarization

Abstract

The volume of information has been increased almost in all fields like education,medical and science etc. This high volume is due to social networking and other activities. The management of data becomes a huge problem. It consumes a lot of time and effort to read the whole information and conclude the results. So the interest in automatic text summarization systems has been increased. It is applied in every field to reduce the information to save time and efforts for analyze the information. Multilingual text summarization is required to provide the summary to user in the desired language. So the all people can take the advantage of information sources which was not possible earlier.The project work has been carried out in multilingual text summarization. The hybrid approach has been used in the system by extracting the best features of existing algorithms; by providing the new features also. The Universal Networking Language (UNL) has been used in system for the language transformation. The natural language document in English is provided as input. The natural language document is converted into UNL. The system produces the summary of the UNL file. The score computation is a major parameter for the sentence selection. The summary is refined in each step by different UNL heuristics. The language transformations are carried out by UNL tools i.e. IAN and EUGENE. At the end summary in Punjabi language is produced.