Combining relations for information extraction from free text

Abstract

Relations between entities of the same semantic type tend to be sparse in free texts. Therefore, combining relations is the key to effective information extraction (IE) on free text datasets with a small set of training samples. Previous approaches to bootstrapping for IE used different types of relations, such as dependency or co-occurrence, and faced the problems of paraphrasing and misalignment of instances. To cope with these problems, we propose a framework that integrates several types of relations. After extracting candidate entities, our framework evaluates relations between them at the phrasal, dependency, semantic frame, and discourse levels. For each of these levels, we build a classifier that outputs a score for relation instances. In order to integrate these scores, we propose three strategies: (1) integrate evaluation scores from each relation classifier; (2) incorporate the elimination of negatively labeled instances in a previous strategy; and (3) add cascading of extracted relations into strategy (2). Our framework improves the state-of-art results for supervised systems by 8%, 15%, 3%, and 5% on MUC4 (terrorism); MUC6 (management succession); ACE RDC 2003 (news, general types); and ACE RDC 2003 (news, specific types) domains respectively.

Learning to Automatically Solve Algebra Word Problems

“…Most work has focused on the supervised case, where the templates are manually defined and data is labeled with alignment information, e.g. (Grishman et al, 2005;Maslennikov and Chua, 2007;Ji and Grishman, 2008;Reichart and Barzilay, 2012). However, some recent work has studied the automatic induction of the set of possible templates from data (Chambers and Jurafsky, 2011;Ritter et al, 2012).…”

Discourse structure and language technology

“…This discourse information is used to filter the wide range of potentially available syntactic paths for linguistic expressions above the clause level (only 2% of which are eventually useful as indicators of semantic relations). Maslennikov and Chua (2007) show that their inclusion of information from discourse structure leads to an improvement of the F-score from 3% to 7% in comparison to other state-of-the-art IE systems that do not take into account discourse structure. However, this strategy basically amounts to reintroducing clause structure into their system because the EDU structures are typically clausal.…”

“…It is fairly straightforward to extract this information after the document has been segmented, as the functional label of a segment is strongly predictive of the information it will contain. Maslennikov and Chua's (2007) approach is different as it assumes a fully hierarchical discourse structure. Their goal is to extract semantic relations between entities, for instance, 'x is located in y'.…”

Information Extraction

“…Maslennikov and Chua (Maslennikov and Chua 2007) use dependency and RST-based discourse relations to connect entities in different clauses and find long-distance dependency relations.…”

“…The tasks in these evaluations were somewhat different, as were the corpora, nevertheless they all seemed to exhibit a ceiling around 60% recall and precision. Although good progress has been made in automating the construction of IE systems using machine learning techniques, current state-of-the-art systems still have not broken through this 60% barrier The 60% barrier FIGURE 21.3: Chronology of MUC system performance in performance on the MUC data sets (e.g., (Soderland 1999;Chieu, Ng, and Lee 2003;Maslennikov and Chua 2007)). …”