Disambiguation of Verbs by Collocation

In this project, we propose using statistical approaches to the analysis of corpus data in order to discover Typical Usage Patterns (TUPs) and hence create a resource for the Disambiguation of Verbs by Collocation (DVC). This project goes beyond the current state-of-the-art represented by word sense disambiguation based on machine-readable dictionaries; typical valency-based approaches, which rarely pay attention to collocations; WordNet, which does not analyse lexical syntagmatics or collocates; and semantic role labelling, which tags mainly thematic roles (e.g. Agent, Patient, Location), rather than semantic types (e.g. Human, Firearm, Route). In our project we propose to associate meanings with normal usage patterns, rather than words in isolation, and to integrate lexical collocations with valency, providing an empirically well-founded resource for use in mapping meaning onto word use in free text. DVC will show the comparative frequency of each pattern of each verb, enabling programs to develop statistically based probabilistic reasoning about meanings, rather than trying to evaluate all possibilities equally.

The internal structure of lexical arguments of verbs will be analysed using computational linguistics techniques, so that for example the relationship between "repair the roof" and "repair the damage" is recognized. Even though the nouns "roof" and "damage" have different semantic types, they activate the same meaning of "repair". Once this has been done, the structural relationship is applied to other verbs, e.g. "treat a patient" and "treat his injuries".

In a pilot project at Masaryk University, Brno, CZ (http://nlp.fi.muni.cz/projects/cpa/), involving analysis of 700 verbs, Prof. Hanks, the co-investigator of the project, showed that, while words may be highly ambiguous, patterns are rarely ambiguous and, furthermore, most uses of most verbs can be assigned unambiguously to a pattern. The existing verbs will be used to train a statistical method the output from which will be verified lexicographically. As the number of annotated verbs increases, the training procedure will be repeated and so improve the accuracy and speed of the annotation. At each step, the researchers employed by the project will analyse the computer output and correct errors. The objective of the DVC project is to analyse 3000 common English verbs and annotate at least 250 corpus lines for each verb. An in-depth data analysis of 100 verbs will be carried out. The resource will be made publicly available at the end of the project.

The DVC project is based on and will contribute to the Theory of Norms and Exploitations (TNE) of Prof. Hanks. TNE says, in essence, that a language consists of two interlinked systems of rules governing word use: a set of rules for the normal uses of words and a second-order set of rules governing the ways in which normal patterns are exploited. Exploitations are deliberately unusual utterances. They play a large role in linguistic change (word-meaning change). as today's exploitation may become tomorrow's norm.

The value of DVC will be proven by textual entailment and paraphrasing, in this way demonstrating its potential usefulness in a large number of fields of computational linguistics which benefit from these two applications.

The project will be disseminated using a wide variety of means. A fully user-friendly publicly available website will contain news about the progress of the project and will provide links to project research papers. It will also host interactive demos that will enable visitors to see the patterns collected and test the technologies developed in the project. Papers will be submitted to international conference and peer-reviewed journals. Evaluation conferences such as SEMEVAL and RTE will be used to assess the methods developed in this project in a standard environment. An important outcome of the project will be a monograph (theory, methodology, empirical findings).

The immediate beneficiaries will be computational linguists working both in academic and commercial environments, who will be able to use DVC as a resource in improving applications such as machine text interpretation, question answering, information retrieval, machine translation, and idiomatic text generation. DVC will provide an inventory of natural, normal and largely unambiguous phraseology (accompanied by meanings for each phrase), with a mechanism for interpreting unusual and imaginative phraseology by "best-match" preferential procedures. For these researchers, it will be a reference resource with a much-needed focus on phraseology and methods for disambiguation, areas of information that are neglected in currently available dictionaries.

It will also provide a resource that can contribute to making a reality of Berners Lee's original dream of the semantic web, in which machines will be able to process the information contained in the natural language of raw text (in contrast to the adding and processing of tags which is the current focus of "semantic web" research).

E-book reading devices such as the Amazon Kindle enable a dictionary to interact with any text that a user is reading. It is technically straightforward to set up a link connecting any word in any electronic text with the relevant entry in an appropriate electronic dictionary. However, for such an application to be fully effective, it is also necessary to enable the software to select, not only the most relevant word, but also the most relevant sense of that word. For this application, currently available research on word-sense relevance is not satisfactory. Research into normal phraseology such as that entailed by the DVC project can be shown to be an essential prerequisite, enabling programmers to match the contexts of word uses in a text with the best-match pattern(s) in a pattern dictionary. By embedding technology like the one developed in the DVC project into e-book readers, it will be possible to enhance the reading experience.

At a theoretical level, DVC results will support developments in empirically well founded theories of language and meaning. It will provide supporting evidence for use by the many researchers who are now beginning to re-evaluate speculative linguistic theories received from a previous generation of linguists. It will do this because it will be based on painstaking analysis of how people actually use words to make meanings in everyday language use, rather than on speculations about boundary-case possibilities. A new generation of linguists is focusing increasingly on the lexicon and aspects such as lexical preferences rather than on deterministic syntax. DVC is a lexical project par excellence, aiming to show how words combine (syntagmatics and collocations) to make meanings.

Researchers into metaphor and figurative language will welcome the many corpus-derived examples of anomalous uses of words, which will be a by-product of DVC's corpus analysis.

DVC will also be of benefit to English language teachers, course-book writers, and lexicographers, showing how words can be grouped into sets and how they combine syntagmatically to make meanings. The Pattern Dictionary of English Verbs, which will be an eventual outcome of DVC research, will aim among other things to identify the normal patterns in which each verb is used. Feedback from the language-teaching community has already indicated that language teachers who are aware of the complexity of corpus data regard such a dictionary as a much-needed resource.