DefinePK

Abstract

Currently there is a growing interest in the automation of extracting the information from natural language text which occurs as a large part of domain knowledge. Software Requirements specification (SRS) enlists all the user’s requirements that can be analyzed through elicitation process from natural language text which has its own limitations. In the present study an attempt is been made to construct a schema for the tables and their inter relationships to all the other tables extracted from the natural language requirements specification text. Initially, the schema for the table is constructed by identification of the Primary Key (PK) attribute based on adjectives, prioritizing the preference of the attributes and hand crafted rules was trained from the statistical data. The Foreign Key (FK) attribute identification is done to construct a relational schema, which establishes the inter relationship between the table attributes from the extracted primary key. Finally, a database which can show the relationship among the tables is built after the identification of the foreign key attributes of a table using highly referenced primary key attribute. By constructing a validated real time automated database, the user can query and acquire domain knowledge.