M.Sc. (Information Technology) / Recommender systems assist a system user to identify relevant content within a specific context. This is typically performed through an analysis of a system user’s rating habits and personal preferences and leveraging these to return one or a number of relevant recommendations. There are numerable contexts in which recommender systems can be applied, such as movies, tourism, books, and music. The need for recommender systems has become increasingly relevant, particularly on the Internet. This is mainly due to the exponential amount of content that is published online on a daily basis. It has thus become more time consuming and difficult to find pertinent information online, leading to information overload. The relevance of a recommender system, therefore, is to assist a system user to overcome the information overload problem by identifying pertinent information on their behalf. There has been much research done within the recommender system field and how such systems can best recommend items to an individual user. However, a growing and more recent research area is how recommender systems can be extended to recommend items to groups, known as group recommendation. The relevance of group recommendation is that many contexts of recommendation apply to both individuals and groups. For example, people often watch movies or visit tourist attractions as part of a group. Group recommendation is an inherently more complex form of recommendation than individual recommendation for a number of reasons. The first reason is that the rating habits and personal preferences of each system user within the group need to be considered. Additionally, these rating habits and personal preferences can be quite heterogeneous in nature. Therefore, group recommendation becomes complex because a satisfactory recommendation needs to be one which meets the preferences of each group member and not just a single group member. The second reason why group recommendation is considered to be more complex than individual recommendation is because a group not only includes multiple personal preferences, but also multiple personality types. This means that a group is more complex from a social perspective. Therefore, a satisfactory group recommendation needs to be one which considers the varying personality types and behaviours of the group. The purpose of this research is to present PerTrust, a generic framework for group recommendation with the purpose of providing a possible solution to the aforementioned issues noted above. The primary focus of PerTrust is how to leverage both personality and trust in overcoming these issues.

published_or_final_version / Computer Science / Master / Master of Philosophy

隨着在線電子商務網站，音樂視頻網站和社會性共享推薦網站的迅速發展，網站用戶面臨爆炸性增長的選擇。前所未見的大量選擇導致信息過載問題。信息過載問題是指由於存在數量巨大的信息，用戶不能有效的理解並做出選擇的問題。推薦系統是解決信息過載問題的一個關鍵組成部分。過去數十年，推薦系統技術有了長足的進步。研究重點又基於臨近用戶的方法向基於模型的方法過度。然而，推薦系統仍然不夠成熟完善。在本論文中，我們基於真實生活中遇到的問題提出改善推薦系統的方法。 / 首先，我們提出推薦系統的在線學習算法。傳統推薦系統使用批量式學習算法進行訓練。這些方法容易理解並且容易實現。然而批量式學習算法不能有效應對當今推薦系統所面臨的動態情況。新的用戶和新的物品不斷加入推薦系統。在批量式學習算法框架下，要將這些新用戶新物品納入系統，需要對所有數據進行重新學習。另外，在批量式學習算法的每一個步驟中，我們需要處理所有的數據。在現今推薦系統規模下，這通常是非常耗時的。在線學習算法可以通過對每一個數據點調整模型而解決上述兩個問題。 / 其次，我們深入調查大量推薦系統所作的一個假設。該假設默認推薦系統蒐集的打分數據的分佈和未蒐集到的打分數據的分佈是完全一致的。我們使用在真實推薦系統中蒐集的數據證明這個假設極不可能爲真。使用失數據理論的方法，我們提出一個不基於改假設的模型。我們的模型放棄了這個假設並且能夠得到公正的推薦。 / 再次，我們詳細調研推薦系統中的垃圾用戶問題。垃圾用戶的打分會污染正常用戶的數據並導致正常用戶的體驗受到影響。我們提出使用用戶聲譽系統去記錄用戶的聲譽並利用用戶的聲譽去區分垃圾用戶和正常用戶。我們提出一個聲譽生成系統的框架。許多聲譽生成系統是我們聲譽生成系統框架的一個實做。基於該框架，我們還提出一個基於矩陣分解的用戶聲譽生成系統。該系統擁有出衆的分辨垃圾用戶的能力。 / 最後，我們將基於內容的推薦和協同過濾推薦有機結合以便減輕乃至解決冷啓動問題。冷啓動問題是指推薦系統中關於某個用戶或物品的信息是如此之少以至於系統不能對該用戶或改物品做出有效的推薦。用戶的文字性評價中通常包含大量用戶喜好和物品屬性信息。但用戶的文字性評價通常都被直接棄。我們提出一個同時使用基於內容的方法去處理用戶文字性評價信息，使用協同過濾方法處理用戶打分的整合式推薦模型。我們的模型能有效減輕冷啓動問題的影響並且對黑盒協同過濾算法提供可理解的詞彙標籤。這些標籤有助於幫助推薦系統提供推薦的原因。 / 綜上所述，在本論文中我們解決了推薦系統面臨的實際問題並從各個方面對傳統推薦系統進行改進。大量真實數據上的實驗驗證我們提出方法的有效性和高效性。 / With the rapid development of e-commerce websites, music and video streaming websites and social sharing websites, users are facing an explosion of choices nowadays. The presence of unprecedentedly large amount of choices leads to the information overload problem, which refers to the difficulty a user faces in understanding an issue and making decisions that are caused by the presence of too much information. Recommender systems learn users’ preferences based on past behaviors and make suggestions for them. These systems are the key component to alleviate and solve the information overload problem. Encouraging progress has been achieved in the research of recommender systems from neighborhood-based methods to model-based methods. However, recommender systems employed today are far from perfect. In this thesis, we propose to improve the recommender systems from four perspectives motivated by real life problems. / First and foremost, we develop online algorithms for collaborative filtering methods, which are widely applicable to recommender systems. Traditionally batch-training algorithms are developed for collaborative filtering methods. They enjoy the advantage of easy to understand and simple to implement. However, the batch-training algorithms fail to consider the dynamic scenario where new users and new items join the system constantly. In order to make recommendations for these new users and on these new items, batchtraining algorithms need to re-train the model from scratch. During the training process of batch-training algorithms, all the data have to be processed in each iteration. This is prohibitively slow given the sheer size of users and items faced by a real recommender system. Online learning algorithms can solve both of the problems by updating the model incrementally based on a rating point. / Secondly, we question an assumption made implicitly by most recommender systems. Most existing recommender systems assume that the rating distribution of collected ratings and that of the unobserved ratings are the same. Using data collected from a real life recommender system, we show that this assumption is unlikely to be true. By employing the powerful missing data theory, we develop a model that drops this unrealistic assumption and makes unbiased predictions. / Thirdly we examine the spam problem confronted by recommender systems. The ratings assigned by spam users contaminate the data of a recommender system and lead to deteriorated experience for normal users. We propose to use a reputation estimation system to keep track of users’ reputations and identify spam users based on their reputations. We develop a unified framework for reputation estimation that subsumes a number of existing reputation estimation methods. Based on the framework, we also develop a matrix factorization based method that demonstrates outstanding discrimination ability. / Lastly, we integrate content-based filtering with collaborative filtering to alleviate the cold-start problem. The cold-start problem refers to the situation where a system has too little information concerning a user or an item to make accurate recommendations. With the readily available rich information embedded in review comments, which are generally discarded, we can alleviate the cold-start problem. Additionally, we can tag the black box collaborative filtering algorithm with interpretable tags that help a recommender system to provide reasons on why items are being recommended. / In summary, we solve some of the major problems faced by recommender systems and improve them from various perspectives in this thesis. Extensive experiments on real life large-scale datasets confirm the effectiveness and efficiency of proposed models. / Detailed summary in vernacular field only. / Detailed summary in vernacular field only. / Detailed summary in vernacular field only. / Detailed summary in vernacular field only. / Detailed summary in vernacular field only. / Detailed summary in vernacular field only. / Ling, Guang. / Thesis (Ph.D.) Chinese University of Hong Kong, 2015. / Includes bibliographical references (leaves 169-184). / Abstracts also in Chinese. / Ling, Guang.

Thesis (M. Phil.)--University of Hong Kong, 2008. / Includes bibliographical references (p. 55-58) Also available in print.

Thesis (M.S.)--Worcester Polytechnic Institute. / Keywords: Data recency problem; Recommender system; Time-based forgetting function; Time-based forgetting strategy; Collaborative filtering system. Includes bibliographical references (p. 73-74).

(1) Relational fusion of multiple features for the classical regression task (single measure and dimension). Originally, the task of recommender systems is formulated as a regression task. Many CF algorithms and fusion methods have been proposed. The limitation of previous fusion methods is that only local features are utilized and the global relational dependency is ignored, which would impair the performance of CF. We propose a relational fusion approach based on conditional random fields (CRF) to improve traditional fusion methods by incorporating global relational dependency. / (2) Fusion of regression-oriented and ranking-oriented algorithms for multi-measure adaption. Beyond the level of classical regression, ranking the items directly is another important task for recommender systems. A good algorithm should adapt to both regression-oriented and ranking-oriented measures. Traditionally, algorithms separately adapt to a single one, thus they cannot adapt to the other. We propose methods to combine them to improve the performances in both measures. / (3) Fusion of quality-based and relevance-based algorithms for multi-dimensional adaption. Recommender systems should consider the performances of multiple dimensions, such as quality and relevance. Traditional algorithms, however, only recommend either high-quality or high-relevance items. But they cannot adapt to the other dimension. We propose both fusion metrics and fusion approaches to effectively combine multiple dimensions for better performance in multi-dimensional recommendations. / (4) Investigation of impression efficiency optimization in recommendation. Besides performance, impression efficiency, which describes how much profit can be obtained per impression of recommendation, is also a very important issue. From recent study, over-quantity recommendation impression is intrusive to users. Thus the impression efficiency should be formulated and optimized. But this issue has rarely been investigated. We formulate the issue under the classical secretary problem framework and extend an online secretary algorithm to solve it. / Recommender systems are important nowadays. With the explosive growth of resources on the Web, users encounter information overload problem. The research issue of recommender systems is a kind of information filtering technique that suggests user-interested items (e.g., movies, books, products, etc.) to solve this problem. Collaborative filtering (CF) is the key approach. Over the decades, recommender systems have been demonstrated important in E-business. Thus designing accurate algorithms for recommender systems has attracted much attention. / This thesis is to investigate effective fusion-based approaches for recommender systems. Effective fusion of various features and algorithms becomes important along with the development of recommendation techniques. Because each feature/algorithm has its own advantages and disadvantages. A combination to get the best performance is desired in applications. The fusion-based approaches investigated are from the following four levels. / Xin, Xin. / Advisers: Wai Lam; Irwin Kuo Chin King; Michael Rung Tsong Lyu. / Source: Dissertation Abstracts International, Volume: 73-06, Section: B, page: . / Thesis (Ph.D.)--Chinese University of Hong Kong, 2011. / Includes bibliographical references (leaves 152-172). / Electronic reproduction. Hong Kong : Chinese University of Hong Kong, [2012] System requirements: Adobe Acrobat Reader. Available via World Wide Web. / Electronic reproduction. [Ann Arbor, MI] : ProQuest Information and Learning, [201-] System requirements: Adobe Acrobat Reader. Available via World Wide Web. / Abstract also in Chinese.

In this thesis, I propose a method for establishing a personalized recommendation system for re-ranking web search results and recommending web contents. The method is based on personal reading interest which can be reflected by the user’s dwell time on each document or webpage. I acquire document-level dwell times via a customized web browser, or a mobile device. To obtain better precision, I also explore the possibility of tracking gaze position and facial expression, from which I can determine the attractiveness of different parts of a document. Inspired by idea of Google Knowledge Graph, I also establish a graph-based ontology to maintain a user profile to describe the user’s personal reading interest. Each node in the graph is a concept, which represents the user’s potential interest on this concept. I also use the dwell time to measure concept-level interest, which can be inferred from document-level user dwell times. The graph is generated based on the Wikipedia. According to the estimated concept-level user interest, my algorithm can estimate a user’s potential dwell time over a new document, based on which personalized webpage re-ranking can be carried out. I compare the rankings produced by my algorithm with rankings generated by popular commercial search engines and a recently proposed personalized ranking algorithm. The results clearly show the superiority of my method. I also use my personalized recommendation framework in other applications. A good example is personalized document summarization. The same knowledge graph is employed to estimate the weight of every word in a document; combining with a traditional document summarization algorithm which focused on text mining, I could generate a personalized summary which emphasize the user’s interest in the document. To deal with images and videos, I present a new image search and ranking algorithm for retrieving unannotated images by collaboratively mining online search results, which consists of online images and text search results. The online image search results are leveraged as reference examples to perform content-based image search over unannotated images. The online text search results are used to estimate individual reference images’ relevance to the search query as not all the online image search results are closely related to the query. Overall, the key contribution of my method lies in its ability to deal with unreliable online image search results through jointly mining visual and textual aspects of online search results. Through such collaborative mining, my algorithm infers the relevance of an online search result image to a text query. Once I estimate a query relevance score for each online image search result, I can selectively use query specific online search result images as reference examples for retrieving and ranking unannotated images. To explore the performance of my algorithm, I tested it both on a standard public image datasets and several modestly sized personal photo collections. I also compared the performance of my method with that of two peer methods. The results are very positive, which indicate that my algorithm is superior to existing content-based image search algorithms for retrieving and ranking unannotated images. Overall, the main advantage of my algorithm comes from its collaborative mining over online search results both in the visual and the textual domains. / published_or_final_version / Computer Science / Doctoral / Doctor of Philosophy

Tracking the evolution of user interests is a problem instance of concept drift learning. Keeping track of multiple interest categories is a natural phenomenon as well as an interesting tracking problem because interests can emerge and diminish at different time frames. The first part of this dissertation presents a Multiple Three-Descriptor Representation (MTDR) algorithm, a novel algorithm for learning concept drift especially built for tracking the dynamics of multiple target concepts in the information filtering domain. The learning process of the algorithm combines the long-term and short-term interest (concept) models in an attempt to benefit from the strength of both models. The MTDR algorithm improves over existing concept drift learning algorithms in the domain. Being able to track multiple target concepts with a few examples poses an even more important and challenging problem because casual users tend to be reluctant to provide the examples needed, and learning from a few labeled data is generally difficult. The second part presents a computational Framework for Extending Incomplete Labeled Data Stream (FEILDS). The system modularly extends the capability of an existing concept drift learner in dealing with incomplete labeled data stream. It expands the learner's original input stream with relevant unlabeled data; the process generates a new stream with improved learnability. FEILDS employs a concept formation system for organizing its input stream into a concept (cluster) hierarchy. The system uses the concept and cluster hierarchy to identify the instance's concept and unlabeled data relevant to a concept. It also adopts the persistence assumption in temporal reasoning for inferring the relevance of concepts. Empirical evaluation indicates that FEILDS is able to improve the performance of existing learners particularly when learning from a stream with a few labeled data. Lastly, a new concept formation algorithm, one of the key components in the FEILDS architecture, is presented. The main idea is to discover intrinsic hierarchical structures regardless of the class distribution and the shape of the input stream. Experimental evaluation shows that the algorithm is relatively robust to input ordering, consistently producing a hierarchy structure of high quality.

As the richness and diversity of information available to us in our everyday lives has expanded, so the need to manage this information grows. The lack of effective information management tools has given rise to what is colloquially known as the information overload problem. Intelligent agent technologies have been explored to develop personalised tools for autonomous information retrieval (IR). However, these so-called adaptive information agents are still primitive in terms of their learning autonomy, inference power, and explanatory capabilities. For instance, users often need to provide large amounts of direct relevance feedback to train the agents before these agents can acquire the users' specific information requirements. Existing information agents are also weak in dealing with the serendipity issue in IR because they cannot infer document relevance with respect to the possibly related IR contexts. This thesis exploits the theories and technologies from the fields of Information Retrieval (IR), Symbolic Artificial Intelligence and Intelligent Agents for the development of the next generation of adaptive information agents to alleviate the problem of information overload. In particular, the fundamental issues such as representation, learning, and classjfication (e.g., classifying documents as relevant or not) pertaining to these agents are examined. The design of the adaptive information agent model stems from a basic intuition in IR. By way of illustration, given the retrieval context involving a science student, and a query "Java", what information items should an intelligent information agent recommend to its user? The agent should recommend documents about "Computer Programming" if it believes that its user is a computer science student and every computer science student needs to learn programming. However, if the agent later discovers that its user is studying "volcanology", and the agent also believes that volcanists are interested in the volcanos in Java, the agent may recommend documents about "Merapi" (a volcano in Java with a recent eruption in 1994). This scenario illustrates that a retrieval context is not only about a set of terms and their frequencies but also the relationships among terms (e.g., java Λ science → computer, computer → programming, java Λ science Λ volcanology → merapi, etc.) In addition, retrieval contexts represented in information agents should be revised in accordance with the changing information requirements of the users. Therefore, to enhance the adaptive and proactive IR behaviour of information agents, an expressive representation language is needed to represent complex retrieval contexts and an effective learning mechanism is required to revise the agents' beliefs about the changing retrieval contexts. Moreover, a sound reasoning mechanism is essential for information agents to infer document relevance with respect to some retrieval contexts to enhance their proactiveness and learning autonomy. The theory of belief revision advocated by Alchourrón, Gärdenfors, and Makinson (AGM) provides a rigorous formal foundation to model evolving retrieval contexts in terms of changing epistemic states in adaptive information agents. The expressive power of the AGM framework allows sufficient details of retrieval contexts to be captured. Moreover, the AGM framework enforces the principles of minimal and consistent belief changes. These principles coincide with the requirements of modelling changing information retrieval contexts. The AGM belief revision logic has a close connection with the Logical Uncertainty Principle which describes the fundamental approach for logic-based IR models. Accordingly, the AGM belief functions are applied to develop the learning components of adaptive information agents. Expectationinference which is characterised by axioms leading to conservatively monotonic IR behaviour plays a significant role in developing the agents' classification components. Because of the direct connection between the AGM belief functions and the expectation inference relations, seamless integration of the information agents' learning and classification components is made possible. Essentially, the learning functions and the classification functions of adaptive information agents are conceptualised by and q d respectively. This conceptualisation can be interpreted as: (1) learning is the process of revising the representation K of a retrieval context with respect to a user's relevance feedback q which can be seen as a refined query; (2) classification is the process of determining the degree of relevance of a document d with respect to the refined query q given the agent's expectation (i.e., beliefs) K about the retrieval context. At the computational level, how to induce epistemic entrenchment which defines the AGM belief functions, and how to implement the AGM belief functions by means of an effective and efficient computational algorithm are among the core research issues addressed. Automated methods of discovering context sensitive term associations such as (computer → programming) and preclusion relations such as (volcanology ⁄→ programming) are explored. In addition, an effective classification method which is underpinned by expectation inference is developed for adaptive information agents. Last but not least, quantitative evaluations, which are based on well-known IR bench-marking processes, are applied to examine the performance of the prototype agent system. The performance of the belief revision based information agent system is compared with that of a vector space based agent system and other adaptive information filtering systems participated in TREC-7. As a whole, encouraging results are obtained from our initial experiments.

Thesis (M.S.)--University of Missouri-Columbia, 2008. / The entire dissertation/thesis text is included in the research.pdf file; the official abstract appears in the short.pdf file (which also appears in the research.pdf); a non-technical general description, or public abstract, appears in the public.pdf file. Title from title screen of research.pdf file (viewed on August 22, 2008) Includes bibliographical references.