Question - Answering System

A Q&A system which can answer various type of questions and provide answers by querying the three databases - Music, Movies and Geography.
The input question is categorized into the respective category and an SQL statement is then framed my parsing the question in top-down approach.
This SQL statement is then used to query the respective database to obtain the answer.

Install Dependencies

• Core NLP Parser
• NLTK
• Wordnet
• SQLite

- from nltk.parse import CoreNLPParser (Runs on localhost:9000)
- pos_tagger = CoreNLPParser(url='http://localhost:9000', tagtype='pos')
- nltk.download('wordnet')
- from nltk.corpus import wordnet
- from nltk.tree import *
- import sqlite3

How to Run the project

1. Run Code: Run the code in command line as:

python <filename>.py "input.txt" "output.txt"

The output in the command line are some results from our implementation The output of the project is written to a textfile named "output.txt"

2. WordNet: This project uses WordNet for categorizing the questions. Some important commands to run before executing:

nltk.download('wordnet')

3. Stanford/CoreNLP: This project uses CoreNLPPaser integrated into nltk

from nltk.parse import CoreNLPParser

Runs on localhost:9000

4. Python Files: Main Script - ssagi2_mkandi3
   Categorization Script - NLP_Part1_Category.py
   WordNet Similarity Script - NLP_Part1_Similarity_WordNet.py
   Movies Category Script - movies_Category.py
   Music Category Script - music_Category.py
   Geography Category Script - geography_Category.py
   Database Script - Quering_On_Database.py
   

Related Documents and Downloads

-> For more details on how to install and CoreNLP API in NLTK, visit: https://github.com/nltk/nltk/wiki/Stanford-CoreNLP-API-in-NLTK
-> Run command to start CoreNLP running:

java -mx4g -cp "*" edu.stanford.nlp.pipeline.StanfordCoreNLPServer \
-preload tokenize,ssplit,pos,lemma,ner,parse,depparse \
-status_port 9000 -port 9000 -timeout 15000 &