No description available.

Factoid Question Answering is an exciting area of Natural Language Engineering that has the potential to replace one major use of search engines today. In this dissertation, I introduce a new method of handling factoid questions whose answers are proper names. The method, Topic Indexing and Retrieval, addresses two issues that prevent current factoid QA system from realising this potential: They can’t satisfy users’ demand for almost immediate answers, and they can’t produce answers based on evidence distributed across a corpus. The first issue arises because the architecture common to QA systems is not easily scaled to heavy use because so much of the work is done on-line: Text retrieved by information retrieval (IR) undergoes expensive and time-consuming answer extraction while the user awaits an answer. If QA systems are to become as heavily used as popular web search engines, this massive process bottle-neck must be overcome. The second issue of how to make use of the distributed evidence in a corpus is relevant when no single passage in the corpus provides sufficient evidence for an answer to a given question. QA systems commonly look for a text span that contains sufficient evidence to both locate and justify an answer. But this will fail in the case of questions that require evidence from more than one passage in the corpus. Topic Indexing and Retrieval method developed in this thesis addresses both these issues for factoid questions with proper name answers by restructuring the corpus in such a way that it enables direct retrieval of answers using off-the-shelf IR. The method has been evaluated on 377 TREC questions with proper name answers and 41 questions that require multiple pieces of evidence from different parts of the TREC AQUAINT corpus. With regards to the first evaluation, scores of 0.340 in Accuracy and 0.395 in Mean Reciprocal Rank (MRR) show that the Topic Indexing and Retrieval performs well for this type of questions. A second evaluation compares performance on a corpus of 41 multi-evidence questions by a question-factoring baseline method that can be used with the standard QA architecture and by my Topic Indexing and Retrieval method. The superior performance of the latter (MRR of 0.454 against 0.341) demonstrates its value in answering such questions.

Vagueness is ubiquitous in natural language. It seems incompatible with classical, bivalent logic, which tells us that every statement is either true or false, and none is vaguely true. Yet we do manage to reason using vague natural language. In fact, the majority of our day-to-day reasoning involves vague terms and concepts. There is a puzzle here: how do we perform this remarkable feat of reasoning? I argue that vagueness is a kind of semantic indecision. In short, that means we cannot say exactly who is bald and who is not because we have never decided the precise meaning of the word 'bald'--there are some borderline cases in the middle, which might be bald or might not. That is a popular general strategy for addressing vagueness. Those who use it, however, do not often say what they mean by 'borderline case'. It is most frequently used in a loose way to refer to in-between items: those people who are neither clearly bald nor clearly not bald. But under that loose description, the notion of borderline cases is ambiguous, and some of its possible meanings create serious problems for semantic theories of vagueness.Here, I clarify the notion of a borderline case, so that borderline cases can be used profitably as a key element in a successful theory of vagueness. After carefully developing my account of borderline cases, I demonstrate its usefulness by proposing a theory of vagueness based upon it. My theory, vagueness as permission, explains how classical logic can be used to model even vague natural language.

No description available.

No description available.

No description available.

No description available.

No description available.

Automatic story generation has a long-standing tradition in the field of Artificial Intelligence. The ability to create stories on demand holds great potential for entertainment and education. For example, modern computer games are becoming more immersive, containing multiple story lines and hundreds of characters. This has substantially increased the amount of work required to produce each game. However, by allowing the game to write its own story line, it can remain engaging to the player whilst shifting the burden of writing away from the game’s developers. In education, intelligent tutoring systems can potentially provide students with instant feedback and suggestions of how to write their own stories. Although several approaches have been introduced in the past (e.g., story grammars, story schema and autonomous agents), they all rely heavily on handwritten resources. Which places severe limitations on its scalability and usage. In this thesis we will motivate a new approach to story generation which takes its inspiration from recent research in Natural Language Generation. Whose result is an interactive data-driven system for the generation of children’s stories. One of the key features of this system is that it is end-to-end, realising the various components of the generation pipeline stochastically. Knowledge relating to the generation and planning of stories is leveraged automatically from corpora and reformulated into new stories to be presented to the user. We will also show that story generation can be viewed as a search task, operating over a large number of stories that can be generated from knowledge inherent in a corpus. Using trainable scoring functions, our system can search the story space using different document level criteria. In this thesis we focus on two of these, namely, coherence and interest. We will also present two major paradigms for generation through search, (a) generate and rank, and (b) genetic algorithms. We show the effects on perceived story interest, fluency and coherence that result from these approaches. In addition, we show how the explicit use of plots induced from the corpus can be used to guide the generation process, providing a heuristically motivated starting point for story search. We motivate extensions to the system and show that additional modules can be used to improve the quality of the generated stories and overall scalability. Finally we highlight the current strengths and limitations of our approach and discuss possible future approaches to this field of research.

Natural language is characterised by enormous linguistic variation (e.g., Fetzer (2003)). Such variation is not random, but is determined by a number of contextual factors. These factors encapsulate the socio-cultural conventions of a speech community and dictate the socially acceptable, i.e. polite, use of language. Producing polite language may not always be a trivial task. The ability to assess a situation with respect to a hearer’s social, cultural or emotional needs constitutes a crucial facet of a speaker’s social and linguistic competence. It is surprising then that it is also a facet which, to date, has received very little attention from researchers in the natural language generation community. Linguistic variation occurs in all linguistic sub-domains including the language of education (Person et al., 1995). Thanks to being relatively more constrained (and hence more predictable with respect to its intentional aspects than normal conversations), teachers’ language is taken in this thesis as a starting point for building a formal, computational model of language generation based on the theory of linguistic politeness. To date, the most formalised theory of linguistic politeness is that by Brown and Levinson (1987), in which face constitutes the central notion. With its two dimensions of Autonomy and Approval, face can be used to characterise different linguistic choices available to speakers in a systematic way. In this thesis, the basic idea of face is applied in the analysis of teachers’ corrective responses produced in real one-to-one and classroom dialogues, and it is redefined to suit the educational context. A computational model of selecting corrective responses is developed which demonstrates how the two dimensions of face can be derived from a situation and how they can be used to classify the many linguistic choices available to teachers. The model is fully implemented using a combination of naive Bayesian Networks and Case-Based Reasoning techniques. The evaluation of the model confirms the validity of the model, by demonstrating that politeness-based natural language generation in the context of teachers’ corrective responses can be used to model linguistic variation and that the resulting language is not singnificantly different from that produced by a human in identical situations.