Skip to content

Boolean Search Engine based on Automobile and Property policy papers.

Notifications You must be signed in to change notification settings

Code-R57/BooleanSearchEngine

Repository files navigation

Boolean Search Engine

The program is a search engine that works on boolean queries and is based on automobile and property policy papers. It also has an option to filter search results by automobile or property tags.
This program is part of an assignment for CS F469 - Information Retrieval course of BITS Pilani, Hyderabad Campus.

Running the Program

  1. Install all the Python libraries required: pip install -r requirements.txt.
  2. Run the GUI using the command: streamlit run GUI.py.
  3. Search for any term and select the appropriate filter to get search results.

NOTE: Please delete the files in inverted_index and affixes directories and run the pre-processsing step to ensure accurate index and presence of all indexed files for the first run.

Operators

Operator Type Operator Example
Boolean AND example AND demo
Boolean OR example OR demo
Boolean ! (NOT) !(example)
Phrase " " "example demo"
Sentence \s exmaple \s demo
Proximity \d exmaple \2 demo
Wildcard * (pre) exm*
Wildcard * (suffix) *ple
Wildcard * (sub) exm*ple
Complex example AND !(demo)

Preprocessing

  1. Add the parsed json file into the respective sub-directory of parsed_json directory.
  2. Run the command: python preprocessing.py or python3 preprocessing.py.

The parsed json file for a document should have the format:

{
    "title" : "demo",
    "sections" : 
    [
        { 
            "section_heading" : "demo",
            "paragraphs":
                [
                    "demo",
                    "demo"
                ]
        },
        {
            "section_heading" : "demo",
            "paragraphs":
                [
                    "demo",
                    "demo"
                ]
        }
            
    ]
}

Group Members

About

Boolean Search Engine based on Automobile and Property policy papers.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages