<?xml version="1.0" standalone="yes"?>
<Paper uid="H05-1013">
  <Title>Proceedings of Human Language Technology Conference and Conference on Empirical Methods in Natural Language Processing (HLT/EMNLP), pages 97-104, Vancouver, October 2005. c(c)2005 Association for Computational Linguistics A Large-Scale Exploration of Effective Global Features for a Joint Entity Detection and Tracking Model</Title>
  <Section position="2" start_page="0" end_page="97" type="intro">
    <SectionTitle>
1 Introduction
</SectionTitle>
    <Paragraph position="0"> In many natural language applications, such as automatic document summarization, machine translation, question answering and information retrieval, it is advantageous to pre-process text documents to identify references to entities. An entity, loosely de ned, is a person, location, organization or geo-political entity (GPE) that exists in the real world.</Paragraph>
    <Paragraph position="1"> Being able to identify references to real-world entities of these types is an important and dif cult natural language processing problem. It involves nding text spans that correspond to an entity, identifying what type of entity it is (person, location, etc.), identifying what type of mention it is (name, nominal, pronoun, etc.) and nally identifying which other mentions in the document it corefers with. The difculty lies in the fact that there are often many ambiguous ways to refer to the same entity. For example, consider the two sentences below: Bill ClintonNAMPER 1 gave a speech today to the SenateNAMORG 2 . The PresidentNOMPER 1 outlined hisPROPER 1 plan for budget reform to themPROORG 2 .</Paragraph>
    <Paragraph position="2"> There are ve entity mentions in these two sentences, each of which is underlined (the corresponding mention type and entity type appear as superscripts and subscripts, respectively, with coreference chains marked in the subscripts), but only two entities: a2 Bill Clinton, The president, his a3 and a2 the Senate, them a3 . The mention detection task is to identify the entity mentions and their types, without regard for the underlying entity sets, while coreference resolution groups a given mentions into sets.</Paragraph>
    <Paragraph position="3"> Current state-of-the-art solutions to this problem split it into two parts: mention detection and coreference (Soon et al., 2001; Ng and Cardie, 2002; Florian et al., 2004). First, a model is run that attempts to identify each mention in a text and assign it a type (person, organization, etc.). Then, one holds these mentions xed and attempts to identify which ones refer to the same entity. This is typically accomplished through some form of clustering, with clustering weights often tuned through some local learning procedure. This pipelining scheme has the signi cant drawback that the mention detection module cannot take advantage of information from the coreference module. Moreover, within the coreference  task, performing learning and clustering as separate tasks makes learning rather ad-hoc.</Paragraph>
    <Paragraph position="4"> In this paper, we build a model that solves the mention detection and coreference problems in a simultaneous, joint manner. By doing so, we are able to obtain an empirically superior system as well as integrate a large collection of features that one cannot consider in the standard pipelined approach.</Paragraph>
    <Paragraph position="5"> Our ability to perform this modeling is based on the Learning as Search Optimization framework, which we review in Section 2. In Section 3, we describe our joint EDT model in terms of the search procedure executed. In Section 4, we describe the features we employ in this model; these include the standard lexical, semantic (WordNet) and string matching features found in most other systems. We additionally consider many other feature types, most interestingly count-based features, which take into account the distribution of entities and mentions (and are not expressible in the binary classi cation method for coreference) and knowledge-based features, which exploit large corpora for learning name-to-nominal references. In Section 5, we present our experimental results. First, we compare our joint system with a pipelined version of the system, and show that joint inference leads to improved performance. Next, we perform an extensive feature comparison experiment to determine which features are most useful for the coreference task, showing that our newly introduced features provide useful new information. We conclude in Section 6.</Paragraph>
  </Section>
class="xml-element"></Paper>