Master of Science / Department of Computing and Information Sciences / Doina Caragea / There is an increasing amount of user-generated information in online documents, includ- ing user opinions on various topics and products such as movies, DVDs, kitchen appliances, etc. To make use of such opinions, it is useful to identify the polarity of the opinion, in other words, to perform sentiment classification. The goal of sentiment classification is to classify a given text/document as either positive, negative or neutral based on the words present in the document. Supervised learning approaches have been successfully used for sentiment classification in domains that are rich in labeled data. Some of these approaches make use of features such as unigrams, bigrams, sentiment words, adjective words, syntax trees (or variations of trees obtained using pruning strategies), etc. However, for some domains the amount of labeled data can be relatively small and we cannot train an accurate classifier using the supervised learning approach. Therefore, it is useful to study domain adaptation techniques that can transfer knowledge from a source domain that has labeled data to a target domain that has little or no labeled data, but a large amount of unlabeled data. We address this problem in the context of product reviews, specifically reviews of movies, DVDs and kitchen appliances. Our approach uses an Adapted Naive Bayes classifier (ANB) on top of the Expectation Maximization (EM) algorithm to predict the sentiment of a sentence. We use grams derived from complete syntax trees or from syntax subtrees as features, when training the ANB classifier. More precisely, we extract grams from syntax trees correspond- ing to sentences in either the source or target domains. To be able to transfer knowledge from source to target, we identify generalized features (grams) using the frequently co-occurring entropy (FCE) method, and represent the source instances using these generalized features. The target instances are represented with all grams occurring in the target, or with a reduced grams set obtained by removing infrequent grams. We experiment with different types of grams in a supervised framework in order to identify the most predictive types of gram, and further use those grams in the domain adaptation framework. Experimental results on several cross-domains task show that domain adaptation approaches that combine source and target data (small amount of labeled and some unlabeled data) can help learn classifiers for the target that are better than those learned from the labeled target data alone.

Master of Science / Department of Computing and Information Sciences / William H. Hsu / Pitch detection and instrument identification can be achieved with relatively high accuracy when considering monophonic signals in music; however, accurately classifying polyphonic signals in music remains an unsolved research problem. Pitch and instrument classification is a subset of Music Information Retrieval (MIR) and automatic music transcription, both having numerous research and real-world applications. Several areas of research are covered in this thesis, including the fast Fourier transform, onset detection, convolution, and filtering. Basic music theory and terms are also presented in order to explain the context and structure of data used. The focus of this thesis is on the representation of musical signals in the frequency domain. Polyphonic signals with many different voices and frequencies can be exceptionally complex. This thesis presents a new model for representing the spectral structure of polyphonic signals: Uniform MAx Gaussian Envelope (UMAGE). The new spectral envelope precisely approximates the distribution of frequency parts in the spectrum while still being resilient to oscillating rapidly (noise) and is able to generalize well without losing the representation of the original spectrum. When subjectively compared to other spectral envelope methods, such as the linear predictive coding envelope method and the cepstrum envelope method, UMAGE is able to model high order polyphonic signals without dropping partials (frequencies present in the signal). In other words, UMAGE is able to model a signal independent of the signal’s periodicity. The performance of UMAGE is evaluated both objectively and subjectively. It is shown that UMAGE is robust at modeling the distribution of frequencies in simple and complex polyphonic signals. Combined classification (combiners), a methodology for learning large concepts, is used to simplify the learning process and boost classification results. The output of each learner is then averaged to get the final result. UMAGE is less accurate when identifying pitches; however, it is able to achieve accuracy in identifying instrument groups on order-10 polyphonic signals (ten voices), which is competitive with the current state of the field.

No description available.

The AEC industry has long been in need of effective modes of information exchange and knowledge sharing, but their practice in the industry is still far from satisfactory. In order to maintain their competence in a highly competitive environment and a globalized market, many organizations in the AEC industry have aimed at a move towards the development of learning organizations. Knowledge management has been seen as an effective way to have every member of an organization engaged in learning at all levels. At the very centre of knowledge management and learning is knowledge sharing through effective communication. Unfortunately, however, there is a big gap in the AEC industry between existing practice and the ideal in this area. In order to effectively coordinate information and knowledge flow in the AEC industry, this present research has developed a framework for an information system – a Construction Information and Knowledge Protocol/Portal (CIKP) which integrates within it a publish/subscribe system, Semantic Web technology, and Social Web concepts. Publish/subscribe is an appropriate many-to-many, people-to-people communication paradigm for handling a highly fragmented industry such as construction. In order to enrich the expressiveness of publications and subscriptions, Semantic Web technology has been incorporated into this system through the development of ontologies as a formal and interoperable form of knowledge representation. This research first involved the development of a domain-level ontology (AR-Onto) to encapsulate knowledge about actors, roles, and their attributes in the AEC industry. AR-Onto was then extended and tailored to create an application-level ontology (CIKP-Onto) which has been used to support the semantics in the CIKP framework. Social Web concepts have been introduced to enrich the description of publications and subscriptions. Our aim has been to break down linear communication through social involvement and encourage a culture of sharing, and in the end, the CIKP framework has been developed to specify desired services in communicating information and knowledge, applicable technical approaches, and more importantly, the functions required to satisfy the needs of a variety of service scenarios.

The History of the Book in Canada / Histoire du livre et de l’imprimé au Canada and other national book history projects have been a catalyst for research into the local production of print and have highlighted the need for more study of the print trades in smaller centres. In Ontario during the nineteenth century independent weekly newspapers were printed in most villages, while larger towns boasted more than one print shop and often one or more booksellers and stationers. Bookbinders were active members of the book trades selling books and stationery, ruling paper, binding local pamphlets, periodicals, and books, and manufacturing blankbooks for a variety of purposes. Since much local printing was ephemeral in nature, the only evidence of its existence is found in the record books kept by printers and binders. Partial business records and other surviving artifacts of the Guelph Bookbindery, which operated from 1855 to 1978, are both a rich source of evidence of the day-to-day operations of the bookbindery and a key to the intersection of print trades in Guelph, Ontario, and the surrounding counties. This study uses local imprints, blankbooks, authors’ papers, newspapers, directories, maps, assessment records, photographs, museum artifacts, and oral history accounts to reconstruct a history of the bookbindery and its place in the print culture of nineteenth-century Guelph. It documents the transformation of a business selling books, stationery, and wallpaper into a commercial bindery, which along with local printers produced large numbers of pamphlets, ubiquitous then but increasingly rare today. It is a view into the microcosm of a dynamic community where print was a vital medium for communication reflecting the cultural, commercial, and entrepreneurial discourse in nineteenth-century Canadian society that reached far beyond its borders.

Web accessibility guidelines have textual representation and provide little support in systematic analysis and usage. The study hypothesizes that the guidelines can be reorganized and represented using the goal oriented technique and design patterns which will allow their usage in a more systematic manner and accommodate detailed analysis of the guidelines with other competing goals. For this study, knowledge from web designers’ experiences in using the guidelines, researchers’ findings, and the actual guidelines have been amalgamated. Six web designers have been consulted on a one-on-one basis and difficulties in using the guidelines for (1) specific scenarios, (2) systematic application, and (3) detailed systems analysis have been reported. Goal oriented modeling and design patterns have been introduced for (1) graphical representation of the guidelines using goal graphs, and (2) keeping the technical details separate from the goal graphs. The proposed representation allows for the accessibility guidelines to be systematically applied into interface design, and systems design using the goal oriented modeling technique.

This research examines situated behaviours and perceptions around textual navigation as it is practiced in situ by professionals working in various domains. In its investigation of interactions between human cognition and mediating artifacts, this research relies heavily on the resources of cognitive ethnography, including both observation and in-depth interviews with participants. Relevant contributions from the fields of information studies, book history, digital humanities, and human-computer interaction are presented to further elucidate the findings of this study. The findings reveal several emergent, interrelated navigational strategies, such as the use of annotations as navigational aids, reliance on automated interface actions, and the navigational value of interface metaphors. In further addressing the practice of textual navigation, this research also describes the creation of a prototype interface reflecting the study’s findings. This research proposes new ways of conceptualizing textual navigation and designing interfaces that support emergent textual interaction.

Heuristic Evaluation (HE) is a popular usability inspection method. Yet little is known about the effect the evaluators have on the outcome of HE. One potentially important feature of evaluators is their end-user status, that is, whether or not they are end-users for whom the interface is designed. I completed a detailed review of the HE literature, combined sources, developed an explicit method for conducting an HE and trained HE novices from different work domains using it. Using these methods I conducted a preliminary randomized crossover study (n=6) of the effect of end-user status during the inspection and merging stages of HE. I estimate a larger study of approximately 148 end-users would be needed to test hypotheses regarding end-user status. I demonstrated a novel measure of the effect of end-user status for the merging stage of HE, which I called the measure of matching similarity (MMS).

Heuristic Evaluation (HE) is a popular usability inspection method. Yet little is known about the effect the evaluators have on the outcome of HE. One potentially important feature of evaluators is their end-user status, that is, whether or not they are end-users for whom the interface is designed. I completed a detailed review of the HE literature, combined sources, developed an explicit method for conducting an HE and trained HE novices from different work domains using it. Using these methods I conducted a preliminary randomized crossover study (n=6) of the effect of end-user status during the inspection and merging stages of HE. I estimate a larger study of approximately 148 end-users would be needed to test hypotheses regarding end-user status. I demonstrated a novel measure of the effect of end-user status for the merging stage of HE, which I called the measure of matching similarity (MMS).

Web accessibility guidelines have textual representation and provide little support in systematic analysis and usage. The study hypothesizes that the guidelines can be reorganized and represented using the goal oriented technique and design patterns which will allow their usage in a more systematic manner and accommodate detailed analysis of the guidelines with other competing goals. For this study, knowledge from web designers’ experiences in using the guidelines, researchers’ findings, and the actual guidelines have been amalgamated. Six web designers have been consulted on a one-on-one basis and difficulties in using the guidelines for (1) specific scenarios, (2) systematic application, and (3) detailed systems analysis have been reported. Goal oriented modeling and design patterns have been introduced for (1) graphical representation of the guidelines using goal graphs, and (2) keeping the technical details separate from the goal graphs. The proposed representation allows for the accessibility guidelines to be systematically applied into interface design, and systems design using the goal oriented modeling technique.