Zoning Ordinance Question Answer

Group Members: Micah Cochran and Seth Lewis

Prerequisites

• Anaconda w/ Python
• Python

Pre-install instructions Debian Linux

From root account:

$ sudo apt-get install git python3-pip python3-venv python3-wheel

Installation

1. Either clone the GitHub Repository:

git clone https://github.com/micahcochran/cs662-qa-land-dev-law-sys.git

OR download the ZIP file from GitHub.

2. Install the conda environment. Installation Details in env/. Note this process may take a while.

Read the README in env/ for next steps in installation.

Command Line Interface

kgqas/cli.py is a command line interface for the Knowledge Graph Question Answering system version.

Folders

• env/ - environment, files to create the environment.

• kgqas/ - Zoning Information Knowledge Graph Question Answering System. Go here to run jupyter notebook (kgqas/KGQAS.ipynb).

• triplesdb/

  • triples knowledge graph "database" is stored here in triplesdb/combined.ttl
  • triplesdb/generate_template.py generates questions for training the models.
  • the .rq files are sample SPARQL queries used during development. Read triplesdb/README.md for more information about running such queries.

• programs/ - see PROGRAMS.md, integration tests and jupyter notebooks. Go here for example code - CURRENTLY NOT WORKING WITH VENV environment.

• proposal/ - This was a proposal of the work before the project. There are some thoughts about how to approach the domain and how we initially thought about approaching the project.

Attribution

Attribution of work:

• nlp/ - corpus loading code was Seth's work, Zoning Ordinance text was Micah's contribution
• kgqas/ - Micah
• programs/ - Seth
• proposal/ - Micah and Seth
• poster/ - Micah's design and Seth contributed his work portions
• tripeldb/ - Micah