Classification is one of the primary tasks of data mining and aims to assign a class label to unseen examples by using a model learned from a training dataset. Most of the accepted classifiers are designed to minimize the error rate but in practice data mining involves costs such as the cost of getting the data, and cost of making an error. Hence the following question arises: Among all the available classification algorithms, and in considering a specific type of data and cost, which is the best algorithm for my problem? It is well known to the machine learning community that there is no single algorithm that performs best for all domains. This observation motivates the need to develop an “algorithm selector” which is the work of automating the process of choosing between different algorithms given a specific domain of application. Thus, this research develops a new meta-learning system for recommending cost-sensitive classification methods. The system is based on the idea of applying machine learning to discover knowledge about the performance of different data mining algorithms. It includes components that repeatedly apply different classification methods on data sets and measuring their performance. The characteristics of the data sets, combined with the algorithm and the performance provide the training examples. A decision tree algorithm is applied on the training examples to induce the knowledge which can then be applied to recommend algorithms for new data sets, and then active learning is used to automate the ability to choose the most informative data set that should enter the learning process. This thesis makes contributions to both the fields of meta-learning, and cost sensitive learning in that it develops a new meta-learning approach for recommending cost-sensitive methods. Although, meta-learning is not new, the task of accelerating the learning process remains an open problem, and the thesis develops a novel active learning strategy based on clustering that gives the learner the ability to choose which data to learn from and accordingly, speed up the meta-learning process. Both the meta-learning system and use of active learning are implemented in the WEKA system and evaluated by applying them on different datasets and comparing the results with existing studies available in the literature. The results show that the meta-learning system developed produces better results than METAL, a well-known meta-learning system and that the use of clustering and active learning has a positive effect on accelerating the meta-learning process, where all tested datasets show a decrement of error rate prediction by 75 %.

This research focuses on cross-cultural differences in facework between English and German conversation. Specifically, the research addresses facework occurring as part and parcel of sociable conversation, as it is played out within moments of focused topic development - what I term 'sociable episodes'. Drawing on extant literature, the study identifies a range of communicative parameters along which English and German communicative style has been shown to differ, non more so than those suggesting different facework norms, and orientation to face needs as opposed to such things as ideational aspects of talk. In an attempt to address these differences, the study develops a model of facework - facework as alignment -which is posited as being appropriate to the study of essentially apolite conversational interaction. Further, although utilising the notions of ritual equilibrium (Goffman 1967) and positive - negative aspects of face (Brown and Levinson), the posited model of facework focuses specifically on aspects of sociable selfhood informing sociable conversation. It is argued that facework in sociable episodes is a matter of positive and negative alignment of sociable selves in and through the claiming of solidarity with and autonomy from other co-participants in terms of expressions of definitions, evaluations, experiences. In terms of English - German differences, these are demonstrated to be a matter of alignment of different sociable selves, ones normatively and routinely positively and negatively aligned in the achievement of sociable conversation, and ones indexing prevailing but culturally differing positive social values (Goffman 1967). The study concludes by identifying areas for future research based on the facework as alignment model developed and applied throughout the thesis.

The acoustic character of a space is often quantified using objective room acoustic parameters. The measurement of these parameters is difficult in occupied conditions and thus measurements are usually performed when the space is un-occupied. This is despite the knowledge that occupancy can impact significantly on the measured parameter value. Within this thesis new methods are developed by which naturalistic signals such as speech and music can be used to perform acoustic parameter measurement. Adoption of naturalistic signals enables passive measurement during orchestral performances and spoken announcements, thus facilitating easy in-situ measurement. Two methods are described within this work; (1) a method utilising artificial neural networks where a network is taught to recognise acoustic parameters from received, reverberated signals and (2) a method based on the maximum likelihood estimation of the decay curve of the room from which parameters are then calculated. (1) The development of the neural network method focuses on a new pre-processor for use with music signals. The pre-processor utilises a narrow band filter bank with centre frequencies chosen based on the equal temperament scale. The success of a machine learning method is linked to the quality of the training data and therefore realistic acoustic simulation algorithms were used to generate a large database of room impulse responses. Room models were defined with realistic randomly generated geometries and surface properties; these models were then used to predict the room impulse responses. (2) In the second approach, a statistical model of the decay of sound in a room was further developed. This model uses a maximum likelihood (ML) framework to yield a number of decay curve estimates from a received reverberant signal. The success of the method depends on a number of stages developed for the algorithm; (a) a pre-processor to select appropriate decay phases for estimation purposes, (b) a rigorous optimisation algorithm to ensure the correct maximum likelihood estimate is found and (c) a method to yield a single optimum decay curve estimate from which the parameters are calculated. The ANN and ML methods were tested using orchestral music and speech signals. The ANN method tended to perform well when estimating the early decay time (EDT), for speech and music signals the error was within the subjective difference limens. However, accuracy was reduced for the reverberation time (Rt) and other parameters. By contrast the ML method performed well for Rt with results for both speech and music within the difference limens for reasonable (<4s) reverberation time. In addition reasonable accuracy was found for EDT, Clarity (C80), Centre time (Ts) and Deutichkeit (D). The ML method is also capable of producing accurate estimates of the binaural parameters Early Lateral Energy Fraction (LEF) and the late lateral strength (LG). A number of real world measurements were carried out in concert halls where the ML accuracy was shown to be sufficient for most parameters. The ML method has the advantage over the ANN method due to its truly blind nature (the ANN method requires a period of learning and is therefore semi-blind). The ML method uses gaps of silence between notes or utterances, when these silence regions are not present the method does not produce an estimate. Accurate estimation requires a long recording (hours of music or many minutes of speech) to ensure that at least some silent regions are present. This thesis shows that, given a sufficiently long recording, accurate estimates of many acoustic parameters can be obtained directly from speech and music. Further extensions to the ML method detailed in this thesis combine the ML estimated decay curve with cepstral methods which detect the locations of early reflections. This improves the accuracy of many of the parameter estimates.

Many English teachers posed the problem that their Arab students were able to construct grammatically correct sentences, but were frequently unable to form them into paragraphs or cohesive texts. In my attest to investigate this problem, I started from the assumption that differing patterns of cohesion in English and Arabic probably account for many difficulties Arab students have in writing English. Sane attempts to look at this, based on a contrastive approach, have already been carried out. For my part, I felt the time had came to look at the systems of Arabic in their own terms, which has not yet been done. For this I followed two avenues of study: Functional Sentence Perspective as developed in the Prague School and Halliday and Hasan's work on textual cohesion. For my purpose I selected four lengthy Arabic texts belonging to different text-types which I first analysed from the Functional Sentence Perspective point of view. For this, I followed Dane's (1974) study of thematic progressions, in order to find out what theme-rheme patterns the different Arabic text-types use. In the next step, I investigated the cohesive ties used in written Arabic Halliday and Hasan's model of textual cohesion (1976). I also compared my texts in order to discover if there is a difference in textual cohesion between text-types in Arabic. My analysis of textual cohesion and text development suggests that: 1. Arabic descriptive texts tend to reiterate the same there in successive sentences. 2. Arabic instructive texts favour the use of the linear thematization of rhemes. 3. Arabic makes inter-clausal relationships explicit. 4. Repetition and parallelism are favoured cohesive devices in all text-types. The thesis consists of an introduction followed by a chapter reviewing various approaches to discourse analysis, a chapter on the text-typological approach which has governed my selection of texts; followed by an account of my methodological approach and my analysis.

This thesis begins by showing the important part that the text plays in disambiguating the temporal interpretaions of the verb forms in Modern Standard Arabic. Proceeding from this point, it shows how, in translation, depending solely on the morphologically-based temporal interpretations of the verb forms or the semantic classifications of the situations represented by the clauses is misleading and cripples the cohesion and coherence of the target text (Chapter One). In order to show the need for this work, the views on the subject of determining the temporal interpretation of the verb-form held by traditional Arab grammarians, Arabists and modern Arab linguists are critically reviewed (Chapter Two). In the body of the thesis (Chapter Three) a method of dealing with this problem is proposed, based on an examination of the text viewed as a whole, not as a combination of decontextualized clauses or forms. Also in this part of the thesis (Chapter Three) a description of the texts analyzed is provided in order to show the authenticity of our data. The approach suggested in chapter three is then given a realistic test by applying it to a fairly large corpus of short news reports produced by and for native speakers of Arabic (Chapter Four). The motivation for this research is the felt need for translators of Arabic and English texts to be made aware of this problem in order to improve the quality of the work they are involved in. The approach that is proposed should also be of benefit to those interested in studying and teaching translation as a process (Chapter Five).

Word order is a textual means available to text producers to arrange their messages in order to convey different communicative functions, the choice of sentence initial position being one of the strategies commonly used to relay pragmatic or textual functions such as emphasis or topic continuity. In Arabic, the use of an initial pre-verbal constituent determines the difference between two types of sentences: the verbal sentence and the nominal sentence. Adopting the Prague School functional approach to grammar as a framework, this study argues that the position of the initial pre-verbal constituent is decided by textual and extra- textual factors. It examines the mechanism of this interaction and its consequences on word order. Arabic word order however, has generally been studied within the transformational-generative framework using sentence-based grammar; even the few cases of functional studies favour this same type of decontextualized data. Very little work of any real significance has been undertaken that attempts to relate sentence type to context and/or text type. In response to this, the data used in this study is text-based and taken from a variety of sources, and the sentences are analyzed in context. Chapter one outlines the aim of the study and reviews current trends in word order research with special regard to Arabic word order, and attempts to highlight the theoretical and empirical limitations of these approaches. Chapter two gives a detailed presentation of the theoretical framework chosen for the analysis. it introduces the corpus used for the analysis, and explains the general method and principles which govern this analysis. Chapter three is the text analysis proper. After a brief presentation of the source text, its general background and organization, each nominal sentence is analyzed textually, contextually, syntactically and functionally. Chapter four discusses the results of the data analysis. The principles at work in word order organization in Arabic are determined, particularly of initial position, and therefore the difference between nominal and verbal sentences. The textual and contextual behaviour of the sentence initial pre-verbal constituent is examined allowing conclusions about the information structure of the Arabic sentence to be drawn.

Effective communication in a foreign language depends on more than knowing the rules of its lexicon, grammar, and phonology. It involves the processing of cultural as well as linguistic knowledge. Any form of communication (and language is one form of communication) has its own strategies. The strategies of language communication vary systematically across languages and cultures. The differences in the general ethos of one community as compared to another lead to differences in the strategies of communication, as certain aspects of the communicative properties of languages might be culture- and language-specific. As a result, foreign language learners might fail to communicate effectively in the foreign language. This failure seems to be greatly due to a transferance of the native-language communicative strategies to the foreign language. The most appropriate way to solve this problem as suggested in this thesis, alongside many other foreign language teachers and applied linguists, is contrastive language work through translation. Such approach makes possible the juxtaposition of the native language and the foreign language, thus allowing the students to see practically and for themselves the culturo-linguistic differences and similarities between the two languages. Otherwise, native language interference will persist and successful performance in the foreign language will not be achieved. The present study sets to investigate two major factors that seem to considerably affect Arab students' learning of English as a foreign language. First, the culturo- educational background which is almost totally ignored as being the second major factor affecting foreign language learning. Second, the linguistic (or mother tongue) factor which, though more researched, no appropriate solutions are yet provided. Due to culturo-educational influence --dealt with in the first part of this thesis-- Arab students tend to approach the foreign language in the same way they approach their native language. They seem to rely on memorization in the study of the foreign language and in their essay-writing. For this reason, they are often accused of plagiarism, a practice for which the blame should not be laid upon them alone. The Arab educational system should undergo- the biggest portion of the blame. Arab students, from a very young age and from the earliest educational stage, the kuttaab, are instructed basically orally and trained to rely heavily on their memories. Though such training suits young children and the major subject they are taught (Qurlan), it does not stop at this stage, nor does it confine itself to that particular subject. Rather, it escorts them up till university level and extends to most subjects. A solution to get over such a problem as to EFL teaching is suggested in the concluding chapter to this thesis. The second major factor affecting foreign language learning --dealt with in the second part of this thesis-- is the influence from the native language, which is seen to manifest itself on the two major levels of language: the micro and the macro levels. The micro level of language is that of the word and sentence. Influence at such a level appears from the early stages of foreign language learning. However, it is not as serious and as persistent as that at the macro level, ie. the discourse or text level. Here, Arab students often make grave deviations from the norms of the foreign language. Such deviations, their nature, and their cultural and linguistic background are discussed through the examination of the major rhetorical and textual characteristics pertaining to Arabic and English. Translation is proposed as an effective approach of teaching composition to advanced EFL students. Such approach, if applied methodically, will help students develop a much needed awareness of the textual pecularities of the, foreign language; an awareness which will sensitize them to the general linguistic differences and, in particular, those of composing that exist between their mother tongue and the foreign language. Besides, it will certainly help them enlarge more quickly and more practically their EFL lexical and idiomatic repertoire.

English is taught as a compulsory subject in general education and some higher education institutions in Sudan. Students are totally dependent on the five to six hours per week of language input provided through formal classroom instruction. Besides limited exposure to the language, there are other factors confounding the teaching and learning of English such as large classes, lack of books, untrained teachers, examination _oriented teaching and learning, and teaching grammarians' grammar. Such factors have contributed to the decline of standards in English to the extent that the pass mark in English has been reduced to 30 percent in the secondary school certificate examination. The students' interlanguage exhibits features indicating heavy reliance on literal translation from Arabic. At least 50% of their errors could be attributed to this interlinguistic transfer, a strategy which is frequently employed due to the lack of the requisite knowledge of the target language. Of all the detrimental factors, the teaching of grammar seems to be the one that is most directly related to the deterioration of the standard in English. It usually takes the form of giving rules, facts and explanations couched in metalinguistic terms, which is at variance with the learners' hypotheses formation process. Reciting rules and facts about the language is the only one thing that untrained teachers can do. Trained teachers also resort to giving rules and facts due to the fact that the situation in the schools and universities is not conducive to developing the language as a skill. Based on the fact that the effectiveness of foreign language teaching in general and the teaching of grammar in particular is greatly reduced when the focus is on giving rules and complicated grammatical analysis, it is the purpose of this study to explore the possibility that the teaching of grammar could profitably be based on the findings of recent studies on interlanguage and learning strategies. The study focusses on the interlinguistic transfer strategy through translation errors in an attempt to arrive at a learnercentred technique of teaching grammar. Based on the analysis of errors, providing students with simple contrastive comparisons between the native and the target language was articipated to be more effective than giving them abstract rules and metalinguistic explanations. The study provides empirical data verifying the effectiveness of simple interlinguistic comparisons in minimizing translation errors. An experiment was conducted in eight secondary schools and the University of Gezira in Wad Medani, Sudan. A total of 714 male and female Arabicspeaking students were pretested, matched and divided into two equal groups in each school. Based on the results of error analysis, two lessons, one normal and one experimental, were developed to teach the relative clauses in English. The normal lesson followed the traditional format of examples, rules and explanations couched in metalinguistic terms. The experimental lesson included terminology-free comparisons of relative clauses in English and Arabic. The two groups were taught by the same teacher in each school and the university. The same pretest was administered as a post-test. The matched group t test was used to compare the means of the active object relative clauses correctly produced by the two groups in each school. A significant difference was observed between the two groups. The experimental group performed better than the normal group. The t values were 6.387 (df=83), 3.240 (df=54), 1.969 (df=29), 1.758 (df=28), 3.043 (df=41), 4.586 (df=35), 2.651 (df=23), 3.030 (df=14), and 3.747 (df=41). The probability that the difference was due to chance was less than 5% in all cases. The findings supported the hypothesis that the error-based interlinguistic comparisons techniques would be more efficient than the currently used traditional technique in minimizing negative transfer errors. The implications of the findings on the teaching of grammar, error correction, materials development and teacher training are discussed together with the limitations of the study and the need for further research to confirm the findings before they can be generalized.

This research is concerned with understanding a quality management system and its interaction with maintenance management activities within the organisation. The interaction between maintenance and quality, can lead to their integration with production. The second output of production is maintenance, whose output is increased production capacity. Both the production process and the quality of the maintenance work, which, in turn, affects equipment condition, affect the quality of final product. The thesis aims to identify the barriers and difficulties affecting quality in maintenance, within Libyan (public sector) manufacturing organisations. The specific objectives are derived to provide focus for the research activities, in order to fulfil the specific aim of the research in a structured and scientific manner. In order to achieve the thesis objectives, an empirically-based systems analysis of two case study organisations in the Libyan public manufacturing sector was carried out. Substantial field work was carried out using predominantly a qualitative approach. Qualitative data was collected by semi structured interview (from different levels of management and supervisors) to explore the quality management phenomena, and to provide a more holistic understanding through triangulation techniques of required and suitable data. This study has contributed to existing knowledge through getting an in-depth understanding of quality and maintenance issues. A specific definition of "quality in maintenance" was developed, common barriers of quality systems and the key factors of improving maintenance were summarised. Furthermore, by using the framework in figure 2.8 the enablers and inhibitors of quality in maintenance were demonstrated. This study is the first one that identified the unique barriers affecting quality in maintenance within the Libyan manufacturing organisations. These barriers are divided into three main kinds: the technical, economic barriers, managerial, organisational barriers, and cultural environmental barriers. The key findings of the research indicate that the case study organisations do not actively promote quality in the maintenance area that takes into consideration the influence of organisational, social, economic and political factors (change culture) on the quality of operations and performance inside the organisations.

The aim of this thesis is to investigate the reasons for the growth of large scale labour migration from Ratnagiri district during the nineteenth century. It is argued firstly that for an understanding of the origins of migration from Ratnagiri it is necessary to investigate the socio-economic structure of the district, since exogenous demand for labour cannot explain many aspects of the pattern of migration from Ratnagiri, nor can it explain the high rate of migration compared to other areas with similar access to labour markets. It is argued that regional and gender patterns of migration from Ratnagiri can be partly explained by the structure of demand for labour within the district; but that the scale of migration can most convincingly be explained in terms of the acute poverty of sections of the rural population. It is argued that this poverty cannot be ascribed to demographic pressure in the early nineteenth century, since population in the district did not rise rapidly until migration was already underway. It is instead suggested that the poverty of many cultivators in the earlier nineteenth century was an outcome of the spread of a village zamindari system in Ratnagiri during the late eighteenth century, the impact of which was intensified by legal changes introduced under British rule; the consequent concentration of landholding in the hands of the village zamindars led to higher exactions on the lower caste cultivators, which stimulated emigration in the mid nineteenth century. Furthermore, it is suggested that the land tenure system was at the root of the problems of agricultural development which the district faced later in the nineteenth century. When population rose In the mid nineteenth century, the extension of cultivation put pressure on the fragile ecology of the district, which led to rapid deforestation and falling yields per acre. it is argued that though cultivation intensified In Ratnagiri during the later nineteenth and early twentieth centuries, the output per head nonetheless probably fell, and the system of land tenure discouraged the adoption of many strategies which might have raised output per head, thus perpetuating the poverty which, it is argued, lay at the root of out-migration from Ratnagiri.