Bill Keller

Detecting a Continuum of Compositionality in Phrasal Verbs

We investigate the use of an automatically acquired thesaurus for measures designed to indicate the compositionality of candidate multiword verbs, specifically English phrasal verbs identified automatically using a robust parser. We examine various measures using the nearest neighbours of the phrasal verb, and in some cases the neighbours of the simplex counterpart and show that some of these correlate significantly with human rankings of compositionality on the test set. We also show that whilst the compositionality judgements correlate with some statistics commonly used for extracting multiwords, the relationship is not as strong as that using the automatically constructed thesaurus.

MaxMax: a graph-based soft clustering algorithm applied to word sense induction

This paper introduces a linear time graph-based soft clustering algorithm. The algorithm applies a simple idea: given a graph, vertex pairs are assigned to the same cluster if either vertex has maximal affinity to the other. Clusters of varying size, shape, and density are found automatically making the algorithm suited to tasks such Word Sense Induction (WSI), where the number of classes is unknown and where class distributions may be skewed. The algorithm is applied to two WSI tasks, obtaining results comparable with those of systems adopting existing, state-of-the-art methods.

Evolutionary induction of stochastic context free grammars

This paper describes an evolutionary approach to the problem of inferring stochastic context-free grammars from finite language samples. The approach employs a distributed, steady-state genetic algorithm, with a fitness function incorporating a prior over the space of possible grammars. Our choice of prior is designed to bias learning towards structurally simpler grammars. Solutions to the inference problem are evolved by optimizing the parameters of a covering grammar for a given language sample. Full details are given of our genetic algorithm (GA) and of our fitness function for grammars. We present the results of a number of experiments in learning grammars for a range of formal languages. Finally we compare the grammars induced using the GA-based approach with those found using the inside-outside algorithm. We find that our approach learns grammars that are both compact and fit the corpus data well.

The Distributional Similarity of Sub-Parses

This work explores computing distributional similarity between sub-parses, i. e., fragments of a parse tree, as an extension to general lexical distributional similarity techniques. In the same way that lexical distributional similarity is used to estimate lexical semantic similarity, we propose using distributional similarity between subparses to estimate the semantic similarity of phrases. Such a technique will allow us to identify paraphrases where the component words are not semantically similar. We demonstrate the potential of the method by applying it to a small number of examples and showing that the paraphrases are more similar than the non-paraphrases.

Formalisms for grammatical knowledge representation

Computational linguists require descriptively powerful, computationally effective formalisms for representing grammatical information or knowledge. A wide variety of formalisms have been employed in natural language processing systems over the past several decades, including simple phrase structure grammars, augmented transition networks, logic grammars and unification-based grammar formalisms. Until fairly recently however, comparatively little attention has been given to the issues which underly good grammar formalism design in computational linguistics.This paper examines a number of fundamental issues in the design of formalisms for representing grammatical knowledge. We begin by examining the role of grammar formalisms in computational linguistics, and the trend towards declarative descriptions of grammar. Grammar formalism design is then considered with respect to choices of linguistic representation and grammar notation. The consequences of some specific design choices for the linguistic and computational utility of grammar formalisms are discussed.

Twitter Paraphrase Identification with Simple Overlap Features and SVMs

We present an approach to identifying Twitter paraphrases using simple lexical over-lap features. The work is part of ongoing re-search into the applicability of knowledge-lean techniques to paraphrase identification. We utilize features based on overlap of word and character n-grams and train support vector machine (SVM). Our results demonstrate that character and word level overlap features in combination can give performance comparable to methods employing more sophisticated NLP processing tools and external resources. We achieve the highest F-score for identifying paraphrases on the Twitter Paraphrase Corpus as part of the SemEval-2015 Task1.

DATR Theories and DATR Models

Evans and Gazdar (Evans and Gazdar, 1989a; Evans and Gazdar, 1989b) introduced DATR as a simple, non-monotonic language for representing natural language lexicons. Although a number of implementations of DATR exist, the full language has until now lacked an explicit, declarative semantics. This paper rectifies the situation by providing a mathematical semantics for DATR. We present a view of DATR as a language for defining certain kins of partial functions by cases. The formal model provides a transparent treatment of DATRs notion of global context. It is shown that DATRs default mechanism can be accounted for by interpreting value descriptors as families of values indexed by paths.

A tractable extension of Linear Indexed Grammars

Vijay-Shanker and Weir (1993) show that Linear Indexed Grammars (LIG) can be processed in polynomial time by exploiting constraints which make possible the extensive use of structure-sharing. This paper describes a formalism that is more powerful than LIG, but which can also be processed in polynomial time using similar techniques. The formalism, which we refer to as Partially Linear PATR (PLPATR) manipulates feature structures rather than stacks.

Children with autism align syntax in natural conversation

Previous experimental work has shown that verbal children with an autism spectrum disorder (ASD) converge linguistically, or align, with an interlocutor, and to the same extent as typical children. However, it is not known whether ASD children align in natural conversation. The studies presented in this paper aimed to address this issue. We measured syntactic alignment in ASD children, first using an experimental task, and second in natural conversation. We found that ASD and typical children aligned to the same extent in both tasks, suggesting that experimental findings about alignment in ASD are ecologically valid. We argue, however, that the experimental measurement of alignment overstates the prevalence of syntactic alignment in childrens conversations.

Managing the policies of non-technical users in a dynamic world

In this paper, we describe the use of description logic as the basis for a policy representation language and show how it is used in our implementation of a policy managed pervasive environment. We compare our approach within this domain to conventional policy management in both implementation and analysis, and highlight the difficulties presented by dynamic environments.