Skip to content

apkp7/Question-Answering-Bot

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

42 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Invoker: Natural Language Question Answering Bot

  • Invoker is an open-domain QA system which answers factoid questions using context-based text mining on multiple Knowledge Bases (KB). It converts the NL question into a parse tree, understands the context and executes SPARQL query to extract answer from a KB.
  • Two techniques have been used - one of them is a grammar rule matching model which uses constituency parse trees and queries the WikiData KB, and the other one is a Machine Learning (ML) model which classifies the questions into 6 main classes employing an Support Vector Machine (SVM) classifier having 82% accuracy, performs syntactic parsing to create an annotated dependency parse tree and queries the DBpedia KB.

Architecuture

Architecture

Dependencies