matchescu-base 0.20.0

Base abstract data types for Entity Resolution

matchescu-base

This Python package includes common abstract data types (adt package), utilities (common package) and generic data extraction algorithms for entity resolution. These abstractions are used in other packages such as:

• matchescu-reference-extraction: extracts entity references from data sources
• matchescu-reference-stores: stores entity references efficiently
• matchescu-comparison-space-generation: generates the comparison space used for matching or clustering
• matchescu-matching: various methods of scoring the similarity of entity references
• matchescu-clustering: various methods of scoring the colocation of entity references
• matchescu-profile-assembly: algorithms used to build concrete entity profiles from specific data structures (tuples, lists or graphs)

On its own, the package may be used to create other structured approaches towards entity resolution, particularly based on the Resolvi reference architecture.

Set up dev environment

1. (optional) install pyenv
2. install Python 3.11
3. install Poetry
4. clone this repository
5. run a couple of shell commands

$ cd <REPO_ROOT>
$ poetry install

Run tests

$ poetry run pytest

Activate virtual environment

$ poetry shell

-or-

$ source .venv/bin/activate