Pranav Anand

The Locus of Ergative Case Assignment: Evidence from Scope

The apparently symmetrical patterns of A/S vs. O and A vs. S/O systems of case opposition often tempt an explanation of ergative as a structural (as opposed to inherent) case. In one class of such implementations, agent and object case are determined by distinct, cross-linguistically universal sources (e.g., T(ense) and v), and the two types of case opposition result from parameters determining whether case on intransitive subjects aligns with objects or agents. This intuition has been formalized both in terms of global case-realization principles within GB – be it via dependence (Marantz 1991) or competition (Bittner & Hale 1996) – and, within the spirit of the minimalist program, in terms of whether A-case or O-case is obligatory (Bobaljik 1993; Laka 1993, 2000). In sum, all of the above proposals agree that the ergative is a structural case differing from nominative only in terms of (a) morphology and (b) whether intransitive subjects align with it. Thus, they all predict that the syntactic behavior of nominative and ergative subjects should be largely parallel. Indeed, to date, no difference in subjecthood properties such as control or binding have been found in “morphologically ergative” languages. Moreover, there is no difference in the A’-status of ergative and nominative subjects (as diagnosed by the nonexistence

Looking under the hood: tools for diagnosing your question answering engine

In this paper we analyze two question answering tasks: the TREC-8 question answering task and a set of reading comprehension exams. First, we show that Q/A systems perform better when there are multiple answer opportunities per question. Next, we analyze common approaches to two subproblems: term overlap for answer sentence identification, and answer typing for short answer extraction. We present general tools for analyzing the strengths and limitations of techniques for these sub-problems. Our results quantify the limitations of both term overlap and answer typing to distinguish between competing answer candidates.

Re-expressing judgment

Abstract Potts (this issue) claims that non-speaker-oriented expressive content may be handled by the introduction of an expressive judge to the context. Given both that expressives may express sentiments temporally and modally displaced from the utterance and that there may be multiple expressive judges per sentence, I argue that a unitary judge is insuffcient, and suggest that non-speaker-oriented expressives are more directly captured as a species of partial quotation.

On the hardness of recognizing triangular line graphs

Given a graph G, its triangular line graph is the graph T (G) with vertex set consisting of the edges of G and adjacencies between edges that are incident in G as well as being within a common triangle. Graphs with a representation as the triangular line graph of some graphG are triangular line graphs, which have been studied under many names including anti-Gallai graphs, 2-in-3 graphs, and link graphs. While closely related to line graphs, triangular line graphs have been dicult to understand and characterize. Van Bang Le asked if recognizing triangular line graphs has an ecient algorithm or is computationally complex. We answer this question by proving that the complexity of recognizing triangular line graphs is NP-complete via a reduction from 3-SAT.

Antecedent Selection for Sluicing: Structure and Content

Sluicing is an elliptical process where the majority of a question can go unpronounced as long as there is a salient antecedent in previous discourse. This paper considers the task of antecedent selection: finding the correct antecedent for a given case of sluicing. We argue that both syntactic and discourse relationships are important in antecedent selection, and we construct linguistically sophisticated features that describe the relevant relationships. We also define features that describe the relation of the content of the antecedent and the sluice type. We develop a linear model which achieves accuracy of 72.4%, a substantial improvement over a strong manually constructed baseline. Feature analysis confirms that both syntactic and discourse features are important in antecedent selection.

Annotating the Implicit Content of Sluices

This paper reports on an eort to develop a linguistically-informed annotation scheme for sluicing (Ross, 1969), ellipsis that leaves behind a wh-phrase. We describe a scheme for annotating the elided content, both in terms of a free text representation and its degree of correspondence with its antecedent. We demonstrate that we can achieve reasonable IAA ( between .78 and .88 across eight annotation types) and describe some of the novel patterns that have arisen from this eort.

PROJECTING AWAY THE CLASS IMBALANCE PROBLEM IN AUTHOR ATTRIBUTION

Author identification algorithms attempt to ascribe document to author, with an eye towards diverse application areas including: forensic evidence, authenticating communications, and intelligence gathering. We view author identification as a single label classification problem, where 2000 authors would imply 2000 possible categories to assign to a post. Experiments with a naive Bayes classifier on a blog author identification task demonstrate a remarkable tendency to over-predict the most prolific authors. Literature search confirms that the class imbalance phenomenon is a challenge for author identification as well as other machine learning tasks. We develop a vector projection method to remove this hazard, and achieve a 63% improvement in accuracy over the baseline on the same task. Our method adds no additional asymptotic computational complexity to naive Bayes, and has no free parameters to set. The projection technique will likely prove useful for other natural language tasks exhibiting class imbalance.

Learning Lexico-Functional Patterns for First-Person Affect

Informal first-person narratives are a unique resource for computational models of everyday events and peoples affective reactions to them. People blogging about their day tend not to explicitly say I am happy. Instead they describe situations from which other humans can readily infer their affective reactions. However current sentiment dictionaries are missing much of the information needed to make similar inferences. We build on recent work that models affect in terms of lexical predicate functions and affect on the predicates arguments. We present a method to learn proxies for these functions from first-person narratives. We construct a novel fine-grained test set, and show that the patterns we learn improve our ability to predict first-person affective reactions to everyday events, from a Stanford sentiment baseline of .67F to .75F.