Improved Pattern Learning for Bootstrapped Entity Extraction

By Sonal Gupta, Christopher D. Manning. 2014

142 citations recorded by the source corpus.

Bootstrapped pattern learning for entity extraction usually starts with seed entities and iteratively learns patterns and entities from unlabeled text. Patterns are scored by their ability to extract more positive entities and less negative entities. A problem is that due to the lack of labeled data, unlabeled entities are either assumed to be negative or are ignored by the existing pattern scoring measures. In this paper, we improve pattern scoring by predicting the labels of unlabeled entities. We use various unsupervised features based on contrasting domain-specific and general text, and exploiting distributional similarity and edit distances to learned entities. Our system outperforms existing pattern scoring algorithms for extracting drug-andtreatment entities from four medical forums.

Subfields: Artificial Intelligence, Natural Language Processing

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Natural Language Processing

Citations in the corpus

Snorkel: Rapid Training Data Creation with Weak Supervision (2017, background): “For example, pattern-based heuristics encompass feature annotations [51] and pattern-bootstrapping approaches [18,20] (Example 2.”
Snorkel: rapid training data creation with weak supervision (2019, background): “For example, pattern-based heuristics encompass feature annotations [64] and pattern-bootstrapping approaches [19,22] (Example 2.”
CoType: Joint Extraction of Typed Entities and Relations with Knowledge Bases (2016, background): “While supervised entity recognition systems [14, 34] focus on a few common entity types, weakly-supervised methods [18, 36] and distantly-supervised methods [41, 54, 26] use large text corpus and a small set of seeds (or a knowledge base) to induce patterns or to train models, and thus can apply to ”
Lightly-supervised Representation Learning with Global Interpretability (2018, methodology): “While semi-supervised learning is not novel (Yarowsky, 1995; Gupta and Manning, 2014), our approach performs better than some modern implementations of these methids such as Gupta and Manning (2014).”
Knowledge Graphs (2020, unknown): “Supervised methods [46, 160, 306] require manually labelling all entity mentions in a training corpus, whereas bootstrapping-based approaches [98, 148, 209, 369] rather require a small set of seed examples of entity mentions from which patterns can be learnt and applied to unlabelled text.”
Training Complex Models with Multi-Task Weak Supervision (2018, methodology): “2017a; Craven and Kumlien 1999; Takamatsu, Sato, and Nakagawa 2012), heuristic patterns (Gupta and Manning 2014; Ratner et al. 2018), feature annotations (Mann and McCallum 2010; Zaidan and Eisner 2008), and noisy crowd labels (Karger, Oh, and Shah 2011; Dawid and Skene 1979).”