Domain specific Information Retrieval System
Problem Statement:
The task is to build a search engine which will cater to the needs of a particular domain. You have to feed your IR model with documents containing information about the chosen domain. It will then process the data and build indexes. Once this is done, the user will give a query as an input. You are supposed to return top 10 relevant documents as the output.
About the project
Dataset used - Kaggle-movie-plots
Have a look at the file Design Architecture. It includes the concepts used along with the modified implementation of the TF-IDF ranking.
Project By:
- Kriti Jethlia: Email- [email protected]
- Jui Pradhan: Email- [email protected]
- Anusha Agarwal: Email- [email protected]
-
Clone the repository : [email protected]:JuiP/Information-Retrieval.git
-
cd Information-Retrieval
-
Run files in the order:
python3 preprocess.py python3 tfidf.py python3 server.py
-
In your browser go to
http://0.0.0.0:3000/
-
Type your query in the search bar and wait till it returns the relevant documents :)
- time
- nltk
- pandas
- pickle
- Numpy
- heapq
- flask
- os