Structured and Logical Representations of Assamese Text for Question-Answering System

摘要

Written documents contain information in a language specific syntax form. Computational processing of such information demands representation in a structured form suitable for handling, processing, and analyzing. Such structured representation of documents enables extraction of knowledge through computational means. Once the textual data are represented in structured form, logical representation also becomes easier. This paper discusses our work on analyzing texts in Assamese language and processing those texts in terms of converting into structured and logical representations. Our emphasis is on the Structured Representation of texts and current study focuses on providing the architecture and processing workflow of the system to output structured form of Assamese text. It also includes system design discussions on how these representations of texts in Assamese language will contribute towards building a Question-Answering system.