Extracting Arabic Relations from the Web

TL;DR

This paper presents a method for extracting Arabic relations from web summaries using minimal user input, achieving moderate to high precision and recall across different relation types.

Contribution

It introduces a novel approach leveraging Google summaries and minimal initial data to automatically extract entity relations in Arabic.

Findings

Precision ranges from 0.61 to 0.75

Recall ranges from 0.71 to 0.83

Best results for (player, club) relation

Abstract

The goal of this research is to extract a large list or table from named entities and relations in a specific domain. A small set of a handful of instance relations is required as input from the user. The system exploits summaries from Google search engine as a source text. These instances are used to extract patterns. The output is a set of new entities and their relations. The results from four experiments show that precision and recall varies according to relation type. Precision ranges from 0.61 to 0.75 while recall ranges from 0.71 to 0.83. The best result is obtained for (player, club) relationship, 0.72 and 0.83 for precision and recall respectively.