WiRe57 : A Fine-Grained Benchmark for Open Information Extraction

Article

William Léchelle, Fabrizio Gotti, Philippe Langlais (2018), WiRe57 : A Fine-Grained Benchmark for Open Information Extraction, arXiv:1809.08962

Open  Information  Extraction  (OIE) systems,  starting with TextRunner, seek to extract all relational tuples expressed in text, without being  bound  to  an  anticipated  list  of  predicates. Such  systems  have  been  used  recently  for  relation extraction, question-answering,  and  for  building domain-targeted knowledge bases, among others.

Subsequent  extractors  (ReVerb,   Ollie,   ClausIE, Stanford Open IE, OpenIE4, MinIE) have sought to improve  yield  and  precision. Despite this, the task definition is underspecified,  and,  to  the  best  of  our knowledge,  there  is no gold standard.

In order to mitigate this problem, we built a reference for the task of Open Information Extraction, on five documents. We tentatively resolve a number of issues that arise, including inference and granularity. We seek to better pinpoint the requirements for the task. We produce our annotation guidelines specifying what is correct to extract and what is not. In turn, we use this reference to score existing Open IE systems. We address the non-trivial problem of evaluating the extractions produced by systems against the reference tuples, and share our evaluation script. Among seven compared extractors, we find the MinIE system to perform best.

Available Resources

The benchmark data and code is available on Github.