Distant Supervision with Imitation Learning

Isabelle Augenstein
University of Sheffield

3:15pm-4:15pm, 30 November 2015
EM G.44

Abstract

Distantly supervised approaches have become popular in recent years as they allow training relation extractors without text-bound annotation, using instead known relations from a knowledge base and a large textual corpus from an appropriate domain. While state of the art distant supervision approaches use off-the-shelf named entity recognition and classification (NERC) systems to identify relation arguments, discrepancies in domain or genre between the data used for NERC training and the intended domain for the relation extractor can lead to low performance. This is particularly problematic for "non-standard" named entities such as "album" which would fall into the MISC category.

We propose to ameliorate this issue by jointly training the named entity classifier and the relation extractor using imitation learning which reduces structured prediction learning to classification learning.

The talk will give an introduction to distant supervision and imitation learning and present experiments from our EMNLP 2015 paper [1].

[1] Isabelle Augenstein, Andreas Vlachos, Diana Maynard (2015). Extracting Relations between Non-Standard Entities using Distant Supervision and Imitation Learning.

Keywords: Imitation learning, Linked Data, Semantic Web, Machine Learning, Natural Language Processing

Bio

Isabelle is a Research Associate in the Sheffield NLP group working on Kalina Bontcheva's Pheme EU project and a fourth year PhD student in the process of finishing her thesis supervised by Fabio Ciravegna and Diana Maynard.

She is interested in researching methods for automatic knowledge extraction, working in the general area of Natural Language Processing (NLP). Her research focusses on methods which do not require any manually annotated training data and instead exploit background information, such as Linked Data. She has an interest in various NLP tasks, among those entity and relation extraction, textual entailment and contradiction detection, computing semantic relatedness and topic detection. For her PhD thesis, she is researching distant supervision for Web information extraction.

Webpage: http://staffwww.dcs.shef.ac.uk/people/I.Augenstein/