The advent of WWW has created a large reservoir of data, including multiple documents with the same theme and concept; As a result, the data becomes overload. It is difficult to meet the complex information needs of the user through a single document related to a specific theme within a given time period. A short summary, which conveys the essence of the document, helps in finding relevant information quickly. But, manual summary creation is tedious task and the existing automatic summary creation methods are also may not guarantee the quality. Therefore, we have developed a low complex automatic summary creation method using machine learning method for summarizing multiple documents. The proposed document summarizer employs Adaptive Artificial Neural Network method for summarizing the documents based on certain relevant features that are essential for categorizing the importance of each sentences. Here, the training speed of ANN is improved with Grasshopper Optimization algorithm by finding the optimal weights of the training network based on training error. The proposed research is carried out in the working platform of JAVA and the results are analyzed with recent research works.
Volume 11 | 10-Special Issue
Pages: 1167-1179
DOI: 10.5373/JARDCS/V11SP10/20192960