<?xml version="1.0" standalone="yes"?>
<Paper uid="W04-2310">
  <Title>Anaphora Resolution in Multi-Person Dialogues</Title>
  <Section position="3" start_page="0" end_page="0" type="intro">
    <SectionTitle>
2 Jepthah
</SectionTitle>
    <Paragraph position="0"> In Jepthah, we adopt an integrated approach towards resolving various different kinds of anaphors occurring in dialogue situations. In this approach we fuse together several heuristics with a new kind of computational linguistic insight - that of the deployment of anaphora chains and we develop a graph-based technique for handling the resolution of various anaphors. An anaphora chain may be described as a referential chain comprising series of mutually co-referential anaphoric elements, generally of more than one type, headed by a referential element.</Paragraph>
    <Paragraph position="1"> The class of anaphors that we aim to resolve is fairly large and includes pronouns, reflexives and deictic anaphors. In terms of distribution, we are dealing with anaphors in subject, object and modifier positions, possessive reflexive, and cataphora. It is may be mentioned here that we deal only with unambiguous cases of plural pronouns, such as both of us, two of you, etc. These are the cases in which the domain of the pronouns is clearly quantified, unlike the case of such instances as all of us or they, etc.</Paragraph>
    <Section position="1" start_page="0" end_page="0" type="sub_section">
      <SectionTitle>
2.1 Graph-theoretic Approach
</SectionTitle>
      <Paragraph position="0"> The entire operation is centered around a graph formulation of the resolution problem in the perspective of the dialogue. We extract all the nouns and pronouns present in the dialogue. Assume there are n nouns and p pronouns in the dialogue. Let the ith noun be represented as Ni, with i [?] n and that Pi represents the ith pronoun, with i [?] p. Now, we construct a graph representation for the problem as follows. Let G be the graph that we are interested in formulating, comprising of a node for every Ni and Pj.Let NGi be the node corresponding to the noun Ni and PGj be the node corresponding to the pronoun Pj.</Paragraph>
      <Paragraph position="1"> Thus, we can split the set of vertices of this graph VG into two parts, the set consisting of NGi , [?]i [?] n and the set consisting of PGj , [?]j [?] p. The set of edges EG for this graph G comprises of two types of directed edges and is constructed as follows. Construct a set of edges E1 which includes a directed edge Ei-j from PGi to NGj , for all pairs PGi and NGj . The other set E2 includes a directed edge Eprimei-j from PGi to PGj for all pair of nodes PGi and PGj such that i negationslash= j. Clearly, we have EG = E1 [?]E2. Let us define a property L on the paths in this graph as follows - a path p satisfies the property L iff it consists of a sequence of edges Ei [?] EG (i [?] length(p)) with exactly one edge Ef from the set E1 and the condition that this is the last edge in the sequence, i.e., Elength(p) [?] Ef.</Paragraph>
      <Paragraph position="2"> Intuitively, this graph represents the set of possible anaphor-antecedent relationships. The set of possible referents of an anaphor represented by the node PGi in the graph G consists of all possible distinct nodes NGk that can be reached from PGi using paths that satisfy the prop-erty L. Let this set be represented as Si. Note here that paths as above of length [?] 2 represent anaphoric chains present in the dialogue. One or more edges in these paths are from one anaphor to another and represent co-reference amongst these anaphors. The antecedent space of an anaphor Pi consists of all nouns and pronouns whose corresponding nodes in the graph G are reachable from PGi by traversing a single edge belonging to EG.</Paragraph>
      <Paragraph position="3"> Now, the idea here is to process this antecedent space and rank all the nodes in Si to determine the most likely antecedent for the anaphor Pi. This ranking is done by attaching weights to the edges present in the graph.</Paragraph>
      <Paragraph position="4"> Every edge is awarded a particular weight (less than 1.0), that is evaluated for every edge using a set of heuristics described in section 2.4. The rank of each node NGk in the set Si is determined by the total weight Wik for that node. Wik is computed as follows - let the weight Wp of each path p be defined as the product of the weights of all the edges lying on that path. Then, Wik is the sum of the weights of all the paths from PGi to NGk , i.e.,summationtextp Wp. Hence, for anaphora resolution, we need to basically design an algorithm or a function to compute the weight for each edge in the graph.</Paragraph>
    </Section>
    <Section position="2" start_page="0" end_page="0" type="sub_section">
      <SectionTitle>
2.2 System Design
</SectionTitle>
      <Paragraph position="0"> The input dialogue is passed to the front end which comprises of the Stanford Serialized Parser and PoS tagger.</Paragraph>
      <Paragraph position="1"> The parser gives the best parse for every input sentence, each of which are then subsequently processed. In the first step we extract all the proper nouns present in the dialogue and initialize our character model base and the graph G that was explained in section 2.1. We then take the sequence of parses corresponding to each subsequent dialogue by a speaker and process them sequentially. Techniques for anaphora resolution are then applied in two phases. In the first phase, a set of constraints is applied to this graph, to prune out edges that represent any unfeasible co-references. In the second phase, a set of heuristics are applied to award weights to edges representing these relationships. After the processing is over and all weights have been obtained, the permissible antecedents for each anaphor are ranked and the most likely antecedent for each is outputted. In case there is a plural anaphor, with quantification over x nouns, the top x likely antecedents are outputted.</Paragraph>
      <Paragraph position="2"> While processing the dialogue, a naive character building is implemented. This is done mainly by focusing on the verbs in the sentences. The subject and object nouns associated with these verbs are selected and their relationship is put in the character model base associated with the speaker of the corresponding dialogue. The system maintains an apriori knowledge base with it containing information like ontology and functionalities of several nouns.</Paragraph>
      <Paragraph position="3"> This combination of apriori and assimilated knowledge is then used to apply certain semantic and pragmatic constraints/heuristics on the graph, as shown in the following sections.</Paragraph>
    </Section>
    <Section position="3" start_page="0" end_page="0" type="sub_section">
      <SectionTitle>
2.3 Constraints
</SectionTitle>
      <Paragraph position="0"> We apply the set of restrictions prior to the set of preferences, thereby narrowing down the candidate set as early as possible. The list of constraints that implement these restrictions in Jepthah are listed as follows - null 1. Deictic Constraint: This is a set of simple con- null straints that are specific to dialogue settings because in such settings we can have the concept of frames of reference with regard to the various speakers involved in the dialogue action.</Paragraph>
      <Paragraph position="1"> 2. Non-coreference (Mitkov, 1999): Syntactic features present in a sentence often lend themselves to be expressed as constraints on anaphora reference. These features are captured by our non-coreference constraints which stipulate that certain pairs of anaphor and noun phrases within the same sentence cannot refer to the same antecedent.</Paragraph>
      <Paragraph position="2">  3. Gender, Number and Person Agreement: This is a low level constraint which requires that anaphors and their antecedents must agree in gender, number and person respectively.</Paragraph>
      <Paragraph position="3"> 4. Constraint on Reflexive Pronoun: A reflexive pronoun such as himself, herself, etc must refer to the subject or the object of the verb in whose clause it lies. In case of ellipsis, however, it may refer to the subject or object of the next higher verb to which the clause is attached.</Paragraph>
      <Paragraph position="4"> 5. Semantic Consistency (Mitkov, 1999): This constraint enforces same semantics of the antecedent as the anaphor under consideration.</Paragraph>
    </Section>
    <Section position="4" start_page="0" end_page="0" type="sub_section">
      <SectionTitle>
2.4 Heuristics
</SectionTitle>
      <Paragraph position="0"> Each preference or heuristic, has a certain weight and awards certain points to every anaphor-antecedent relationship. These points are a measure of the likelihood of that anaphor-antecedent relationship. The weight of an edge is the sum total of the weights awarded by each individual heuristic to the anaphor-antecedent relationship.</Paragraph>
      <Paragraph position="1"> The heuristics used in our system are enumerated as fol- null lows 1. Definiteness (Lappin and Leass, 1994): According to this heuristic, nouns that are preceded by a demonstrative pronoun or a definite article are more likely to be antecedents and are awarded higher credibilities.</Paragraph>
      <Paragraph position="2"> 2. Non-prepositional NP (Lappin and Leass, 1994):  This heuristic states that a noun phrase which occurs within a prepositional phrase is less probable to be an antecedent to an anaphor and consequently, it is  awarded less credibility.</Paragraph>
      <Paragraph position="3"> 3. Pleonastic (Lappin and Leass, 1994): This heuristic is based on the observation that there exist some syntactic patterns such that every it anaphor occurring in any of those patterns must be pleonastic. 4. Syntactic Parallelism (Lappin and Leass, 1994):  As per this heuristic, preference is given to noun phrases with the same syntactic function as the anaphor.</Paragraph>
      <Paragraph position="4"> 5. Recency (Mitkov, 1999): This is a very simple heuristic according to which, everything else being comparable, a higher credibility is awarded to the antecedent nearer to the anaphor.</Paragraph>
    </Section>
  </Section>
class="xml-element"></Paper>