NLIIRS: A Question and Answering System for Unstructured Information Using Annotated Text Segment Comparison

Abstract

In the current scenario, handling of the unstructured data and extracting information is one of the most crucial aspects. This paper proposes a novel system, Natural Language Information Interpretation and Representation System (NLIIRS) which can accept information in natural language text and can answer to the queries without storing or converting the data into natural language text. Nowadays the presence of available information is large in number in the light of unstructured data. The unstructured information that we get is in the shape of natural languages texts. Protection is needed for the government operative information or any delicate data so that it might be best used when the data can be extricated effectively and effortlessly. Natural Language Information Interpretation and Representation System (NLIIRS) acknowledges the data as characteristic natural language text, process the data and enables the client to recover data by rendering question in natural language. The inquiries subsequently asked are reacted by NLIIRS as expression based answers. The entire frameworks have been composed on Natural Language Tool Kit (NLTK) of Stanford University which helped us to create POS tag, tokenize the information, and shaping the tree structure. The Noval content handling calculation uses the lemmatizer, stemmer and ne chunker to set up the content for data recovery through Q&A. The upside of this framework is that it need not bother with preparing or training. This framework will empower the client to recover any data of his/her decision from the accessible unstructured data.