published_or_final_version / Civil Engineering / Master / Master of Philosophy

published_or_final_version / Education / Master / Master of Education

published_or_final_version / Statistics and Actuarial Science / Doctoral / Doctor of Philosophy

published_or_final_version / Physics / Doctoral / Doctor of Philosophy

published_or_final_version / Community Medicine / Master / Master of Public Health

published_or_final_version / Industrial and Manufacturing Systems Engineering / Doctoral / Doctor of Philosophy

Any guidance that might help to reduce the problems of accurately estimating software development effort could assist software producers to set more realistic budgets for software projects. This investigation attempted to make a contribution to this by documenting some of the practical problems with introducing structured effort estimation models at a site in the United Kingdom of an international supplier of telephone switching software. The theory of effort modelling was compared with actual practice by examining how the estimating experts at the telephone switching software producer currently carried out estimating. Two elements of the estimation problem emerged: judging the size of the job to be done and gauging the productivity of the development environment. Expert opinion was particularly important to the initial process, particularly when existing software was being enhanced. The study then identified development effort drivers and customised effort models applicable to real-time telecommunications applications. Many practical difficulties were found concerning the actual methods used to record past project data, although the issues surrounding these protocols appeared to be rarely dealt with explicitly in the research literature. The effectiveness of the models was trialled by forecasting the effort for some new projects and then comparing these estimates with the actual effort. The key research outcomes were, firstly the identification and validation of a set of relevant functional effort drivers applicable in a real-time telecommunications software development environment and the building of an effective effort model, and, secondly, the evaluation of alternative prediction approaches including analogy or case-based reasoning. While analogy was a useful tool, some methods of implementing analogy were flawed theoretically and did not consistently outperform 'traditional' model building techniques such as Least Squares Regression (LSR) in the environment under study. This study would, however, support analogy as a complementary technique to algorithmic modelling

This thesis is concerned with the use of models in chemical education, in particular the way chemistry students use models in understanding chemistry. The study begins with an outline of the reasons for my interest in the subject of models in chemistry. The report describes some of the problems encountered by undergraduate chemistry students in dealing with three dimensional structures, and surveys literature relating to visualization skills in chemistry. Preliminary broad areas for investigation are identified including the relevance of models to students, problems of representing 3D structures, and students' use of models to solve chemical problems. A pilot study to gather information in these areas and to develop a suitable research instrument for investigation is described. The pilot study proved useful in highlighting errors in understanding chemical concepts, assessing practical model use, and considering students' perceptions of the relevance of models. This is followed by a consideration of the role of models in understanding chemistry in relation to the nature of model, chemistry and understanding, and the links and interactions between them. It discusses issues such as the match between the currently perceived roles of models in chemical practice and in chemistry teaching and the desirability of bringing these into line. It surveys the literature concerning purpose of model use in chemistry and describes the features of chemical structure models used in the research. Criteria for selecting appropriate data collection and analysis methods in a research study are considered and 80me of the methods adopted in recent chemical education research described. The chapter concludes with a description and justification of the particular research methods used in the study. The report gives details of the interviews carried out with selected scientists to consider the notion of 'the good chemist'. It then describes the videorecorded workshop interviews with forty five chemistry students relating to their appreciation and use of models in chemistry, and the follow up int.erviews with eight of the participants. Data from these interviews are analysed in an attempt to answer the research questions posed initially, including individual chemists' purposes in using models, patterns in model appreciation, perceptions of the good chemist, fiexibility of model use as an indicator of competence as a chemist, and the potential of the workshop interview in higher education assessment. The research findings are discussed in relation to existing literature, and the study concludes with a discussion of the implications for chemistry curricula, chemical education and for future research.

This thesis describes a model of speech timing, predicting at the syllable level, with sensitivity to rhythmic factors at the foot level, that predicts segmental durations by a process of accommodation into the higher-level timing framework. The model is based on analyses of two large databases of British English speech; one illustrating the range of prosodic variation in the language, the other illustrating segmental duration characteristics in various phonetic environments. Designed for a speech synthesis application, the model also has relevance to linguistic and phonetic theory, and shows that phonological specification of prosodic variation is independent of the phonetic realisation of segmental duration. It also shows, using normalisation of phone-specific timing characteristics, that lengthening of segments within the syllable is of three kinds: prominence-related, applying more to onset segments; boundary-related, applying more to coda segments; and rhythm/rate-related, being more uniform across all component segments. In this model, durations are first predicted at the level of the syllable from consideration of the number of component segments, the nature of the rhyme, and the three types of lengthening. The segmental durations are then constrained to sum to this value by determining an appropriate uniform quantile of their individual distributions. Segmental distributions define the range of likely durations each might show under a given set of conditions; their parameters are predicted from broad-class features of place and manner of articulation, factored for position in the syllable, clustering, stress, and finality. Two parameters determine the segmental duration . pdfs, assuming a Gamma distribution, and one parameter determines the quantile within that pdf to predict the duration of any segment in a given prosodic context. In experimental tests, each level produced durations that closely fitted the data of four speakers of British English, and showed performance rates higher than a comparable model predicting exclusively at the level of the segment.

This thesis deals with topic model evaluation and new topic detection in microblogs. Microblogs are short and thus may not carry any contextual clues. Hence it becomes challenging to apply traditional natural language processing algorithms on such data. Graphical models have been traditionally used for topic discovery and text clustering on sets of text-based documents. Their unsupervised nature allows topic models to be trained easily on datasets meant for specific domains. However the advantage of not requiring annotated data comes with a drawback with respect to evaluation difficulties. The problem aggravates when the data comprises microblogs which are unstructured and noisy. We demonstrate the application of three types of such models to microblogs - the Latent Dirichlet Allocation, the Author-Topic and the Author-Recipient-Topic model. We extensively evaluate these models under different settings, and our results show that the Author-Recipient-Topic model extracts the most coherent topics. We also addressed the problem of topic modeling on short text by using clustering techniques. This technique helps in boosting the performance of our models. Topical alignment is used for large scale assessment of topical relevance by comparing topics to manually generated domain specific concepts. In this thesis we use this idea to evaluate topic models by measuring misalignments between topics. Our study on comparing topic models reveals interesting traits about Twitter messages, users and their interactions and establishes that joint modeling on author-recipient pairs and on the content of tweet leads to qualitatively better topic discovery. This thesis gives a new direction to the well known problem of topic discovery in microblogs. Trend prediction or topic discovery for microblogs is an extensive research area. We propose the idea of using topical alignment to detect new topics by comparing topics from the current week to those of the previous week. We measure correspondence between a set of topics from the current week and a set of topics from the previous week to quantify five types of misalignments: \textit{junk, fused, missing} and \textit{repeated}. Our analysis compares three types of topic models under different settings and demonstrates how our framework can detect new topics from topical misalignments. In particular so-called \textit{junk} topics are more likely to be new topics and the \textit{missing} topics are likely to have died or die out. To get more insights into the nature of microblogs we apply topical alignment to hashtags. Comparing topics to hashtags enables us to make interesting inferences about Twitter messages and their content. Our study revealed that although a very small proportion of Twitter messages explicitly contain hashtags, the proportion of tweets that discuss topics related to hashtags is much higher. / text