Natural Language Search Over Database
Aritra DattaGupta · TRAP@NCI (National College of Ireland) · 2016
Natural language search from a database makes it possible for non technical users to fetch information from database in natural language. In natural language parsing getting highly accurate syntactic analysis is highly important. Parsing of natural language means mapping natural language components to its syntactic representation. This system proposes a model to implement this natural language parsing using NLTK chunking with regular expression called \\grammar" and named entity recognition to extract keywords and phrases from the parse tree. The extracted keywords are then mapped to different attributes and commands of a SQL query using an algorithm. The algorithm uses rules to map different commands and attributes of the database. When the algorithm finishes forming the SQL query it executes it into the database and fetches the result to display the information to the user.