Grammatical Case Based IS-A Relation Extraction with Boosting for Polish

TL;DR

This paper introduces a novel method for extracting IS-A relations from Polish text using grammatical case and morpho-syntactic annotations, along with a boosting technique to enhance extraction coverage, tested on a large web corpus.

Contribution

It presents a new approach leveraging grammatical case for IS-A relation extraction and introduces pseudo-subclass boosting to improve relation extraction recall.

Findings

Effective extraction of IS-A relations from Polish web data.

Boosting increases the number of extracted relations.

Method outperforms traditional pattern-based approaches.

Abstract

Pattern-based methods of IS-A relation extraction rely heavily on so called Hearst patterns. These are ways of expressing instance enumerations of a class in natural language. While these lexico-syntactic patterns prove quite useful, they may not capture all taxonomical relations expressed in text. Therefore in this paper we describe a novel method of IS-A relation extraction from patterns, which uses morpho-syntactical annotations along with grammatical case of noun phrases that constitute entities participating in IS-A relation. We also describe a method for increasing the number of extracted relations that we call pseudo-subclass boosting which has potential application in any pattern-based relation extraction method. Experiments were conducted on a corpus of about 0.5 billion web documents in Polish language.