Thesis (M.S.)--State University of New York at Binghamton, Thomas J. Watson School of Engineering and Applied Science, Department of Computer Science, 2008. / Includes bibliographical references.

Thesis (MCIS)--AUT University, 2009. / Includes bibliographical references. Also held in print ( leaves : ill. ; 30 cm.) in the Archive at the City Campus (T 006.312 CRI)

Thesis (Ph. D.)--Oregon State University, 2008. / Printout. Includes bibliographical references. Also available on the World Wide Web.

Recommender systems (RS) have become increasingly popular in many web applications for eliminating online information overload and making personalized suggestions to users. In recent years, user personality has been recognized as valuable info to build more personalized recommender systems. However, the existing personality-based recommender systems has mainly focused on revealing the impact of personality on the user's preference over a single item or an attribute, which may ignore the impact of personality on users' perceptions of recommender systems when multiple recommendations are returned at the same time. In addition, they have mostly relied on personality quiz to explicitly acquire users' personality, which unavoidably demands user efforts. From users' perspective, they may be unwilling to answer the quiz for the sake of saving efforts or protecting their privacy. The application of existing personality-based recommender systems will thus be limited in real life.;In this thesis, we aim at 1) incorporating personality into top-N (N > 1) recommendations, with emphases on personalizing recommendation diversity and improving the recommendation interface design, 2) deriving users' personality from their implicit behavior for augmenting the existing recommender systems.;Specifically, we first develop a generalized, dynamic diversity adjusting approach based on user personality with the goal of achieving personalized diversity tailored to individual users' intrinsic needs. In particular, personality is integrated into a greedy re-ranking process, by which we select the item that can best balance accuracy and personalized diversity at each step, and then produce the final recommendation list. In this approach, personality is both used to estimate each user's diversity preference and to alleviate the cold-start problem of collaborative filtering recommendations. The experimental results demonstrate that our personalized diversity-oriented approach significantly outperforms related methods (including both non-diversity-oriented and diversity-oriented methods) in terms of both accuracy and diversity metrics, especially in the cold-start setting.;In addition to the algorithm development, designing diversity-oriented interface has been proven helpful to augment users' perception of recommendation diversity. However, little work has been done to identify the impact of users' personality on their preference for different types of recommendation interfaces (e.g., the diversity-oriented interface and the non-diversity-oriented interface). In order to fill the gap, we conduct a within-subject user study. We concretely compare a diversity-oriented organization-based recommendation interface with the standard ranked list interface covering three product domains with different investment levels and users' purchase experiences (i.e., mobile phone, hotel and movie). We find that users' perceptions of different recommendation interface are influenced by the product types. More notably, we identify the important role of users' personality in influencing their preference for recommendation interfaces. For instance, introverted users tend to reuse the organization-based interface in the future than the standard ranked list. The results can hence be constructive for improving existing recommendation interface design by considering users' personality.;Although personality has been proven effective at enhancing the multiple recommendations, the effort of explicitly acquiring users' personality traits via psychological questionnaire is unavoidably high, which may impede the application of personality-based recommenders in real life. We hence propose a generalized method to derive users' personality from their implicit behavior and further improve the existing recommender systems. A preliminary experiment has been conducted in movie domain. More specifically, we first identify a set of behavioral features through experimental validation, and develop inference model based on Gaussian Process to unify these features for determining users' big-five personality traits. We then test the model in a collaborative filtering based recommending framework on two real-life movie datasets, which demonstrates that our implicit personality based recommending algorithm significantly outperforms related methods in terms of both rating prediction and ranking accuracy. The experimental results point out an effective solution to boost the applicability of personality-based recommender systems in online environment.

By now, people are accustomed to getting some personalized recommendations when they are finding movies to watch, music to listen, and so on. All of these recommendations come from recommender systems, and can aid the process of the decision making to avoid the problem of "information overload". Over the years, there has been much work done both in industry and academia on developing new approaches for recommender systems. However, there are still some hurdles in adapting recommender systems to a broader range of real-life applications. In the e-commerce environment especially with the so called high-risk products (also called high-cost or high-involvement products, such as digital cameras, computers, and cars), because a user does not buy the high-risk product very often, it is normal that s/he is not able to rate many products. For the same reason, the current buyer is often a new user because s/he would not afford to buy the same kind of high-risk product before. The traditional recommender techniques (such as user-based collaborative filtering and content-based methods) can thus not be effectively applicable in this environment, because they largely assume that the users have prior experiences with products. Thus, the "data sparsity" and "new users" are two typical challenging issues that the classical recommender systems cannot well address in high-risk product domains. In some recommender systems, a new user will be asked to indicate his/her preferences on some aspects in order to address the so called cold-start problem via collecting some preferences. Such collected preferences are usually not complete due to unfamilaring with the product domain, which are called partial preferences.;In this thesis, we propose to leverage some auxiliary data of online reviewers' opinions, so as to enrich the partial preferences. With the objective of developing more effective recommender systems for high-risk products in e-commerce, in our work, we have exerted to derive reviewers' preferences from the textual reviews they posted. Then, these recovered preferences are leveraged to estimate and supplement a new buyer's preference with which the product recommendation is produced. Firstly, we propose a novel clustering method based on Latent Class Regression model (LCRM), which is able to consider both the overall ratings and feature-level opinion values (as extracted from textual reviews) to infer individual reviewers' weight feature preferences, that represent the weights the user places on different product features. Secondly, we propose a method to estimate reviewers' value preferences (i.e., the user's preferences on the product's attribute values) by matching their review opinions to the corresponding attributes' static specifications. Thirdly, we investigate how to combine weight preferences and value preferences to model user preferences based on Multi-Attribute Utility Theory (MAUT) with the purpose of providing higher quality product recommendations. Particularly, it was shown from our experimental studies that the incorporation of review information can significantly enhance the recommendation accuracy, relative to those without considering reviews. As the practical implication, our proposed solutions can be usefully plugged into an online system to be adopted in real-ecommerce sites.

M. Tech. Information Networks / The growth of the internet has made it easy for people to exchange information resulting in the abundance of information commonly referred to as information overload. It causes retailers to fail to make adequate sales since the customers are swamped with a lot of options and choices. To lessen this problem retailers have begun to find it useful to make use of algorithmic approaches to determine which content to show consumers. These algorithmic approaches are known as recommender systems. Collaborative Filtering recommender systems suggest items to users based on other users reported prior experience with those items. These systems are, however, vulnerable to shilling attacks since they are highly dependent on outside sources of information. Shilling is a process in which syndicating users can connive to promote or demote a certain item, where malicious users benefit from introducing biased ratings. It is, however, critical that shilling detection systems are implemented to detect, warn and shut down shilling attacks within minutes. Modern patented shilling detection systems employ: (a) classification methods, (b) statistical methods, and (c) rules and threshold values defined by shilling detection analysts, using their knowledge of valid shilling cases and the false alarm rate as guidance. The goal of this dissertation is to determine a context for, and assess the performance of Meta-Learning techniques that can be integrated in the shilling detection process.

Wang, Dingyan. / Thesis (M.Phil.)--Chinese University of Hong Kong, 2011. / Includes bibliographical references (p. 97-110). / Abstracts in English and Chinese. / Abstract --- p.ii / Acknowledgement --- p.vi / Chapter 1 --- Introduction --- p.1 / Chapter 1.1 --- Overview --- p.1 / Chapter 1.2 --- Motivations --- p.6 / Chapter 1.3 --- Contributions --- p.9 / Chapter 1.4 --- Organizations of This Thesis --- p.11 / Chapter 2 --- Background --- p.14 / Chapter 2.1 --- Label Propagation Learning Framework --- p.14 / Chapter 2.1.1 --- Graph-based Semi-supervised Learning --- p.14 / Chapter 2.1.2 --- Green's Function Learning Framework --- p.16 / Chapter 2.2 --- Recommendation Methods --- p.19 / Chapter 2.2.1 --- Traditional Memory-based Methods --- p.19 / Chapter 2.2.2 --- Traditional Model-based Methods --- p.20 / Chapter 2.2.3 --- Label Propagation Recommendation Models --- p.22 / Chapter 2.2.4 --- Latent Feature Recommendation Models . --- p.24 / Chapter 2.2.5 --- Social Recommendation Models --- p.25 / Chapter 2.2.6 --- Tag-based Recommendation Models --- p.25 / Chapter 3 --- Recommendation with Latent Features --- p.28 / Chapter 3.1 --- Motivation and Contributions --- p.28 / Chapter 3.2 --- Item Graph --- p.30 / Chapter 3.2.1 --- Item Graph Definition --- p.30 / Chapter 3.2.2 --- Item Graph Construction --- p.31 / Chapter 3.3 --- Label Propagation Recommendation Model with Latent Features --- p.33 / Chapter 3.3.1 --- Latent Feature Analysis --- p.33 / Chapter 3.3.2 --- Probabilistic Matrix Factorization --- p.35 / Chapter 3.3.3 --- Similarity Consistency Between Global and Local Views (SCGL) --- p.39 / Chapter 3.3.4 --- Item-based Green's Function Recommendation Based on SCGL --- p.41 / Chapter 3.4 --- Experiments --- p.41 / Chapter 3.4.1 --- Dataset --- p.43 / Chapter 3.4.2 --- Baseline Methods --- p.43 / Chapter 3.4.3 --- Metrics --- p.45 / Chapter 3.4.4 --- Experimental Procedure --- p.45 / Chapter 3.4.5 --- Impact of Weight Parameter u --- p.46 / Chapter 3.4.6 --- Performance Comparison --- p.48 / Chapter 3.5 --- Summary --- p.50 / Chapter 4 --- Recommendation with Social Network --- p.51 / Chapter 4.1 --- Limitation and Contributions --- p.51 / Chapter 4.2 --- A Social Recommendation Framework --- p.55 / Chapter 4.2.1 --- Social Network --- p.55 / Chapter 4.2.2 --- User Graph --- p.57 / Chapter 4.2.3 --- Social-User Graph --- p.59 / Chapter 4.3 --- Experimental Analysis --- p.60 / Chapter 4.3.1 --- Dataset --- p.61 / Chapter 4.3.2 --- Metrics --- p.63 / Chapter 4.3.3 --- Experiment Setting --- p.64 / Chapter 4.3.4 --- Impact of Control Parameter u --- p.65 / Chapter 4.3.5 --- Performance Comparison --- p.67 / Chapter 4.4 --- Summary --- p.69 / Chapter 5 --- Recommendation with Tags --- p.71 / Chapter 5.1 --- Limitation and Contributions --- p.71 / Chapter 5.2 --- Tag-Based User Modeling --- p.75 / Chapter 5.2.1 --- Tag Preference --- p.75 / Chapter 5.2.2 --- Tag Relevance --- p.78 / Chapter 5.2.3 --- User Interest Similarity --- p.80 / Chapter 5.3 --- Tag-Based Label Propagation Recommendation --- p.83 / Chapter 5.4 --- Experimental Analysis --- p.84 / Chapter 5.4.1 --- Douban Dataset --- p.85 / Chapter 5.4.2 --- Experiment Setting --- p.86 / Chapter 5.4.3 --- Metrics --- p.87 / Chapter 5.4.4 --- Impact of Tag and Rating --- p.88 / Chapter 5.4.5 --- Performance Comparison --- p.90 / Chapter 5.5 --- Summary --- p.92 / Chapter 6 --- Conclusions and Future Work --- p.94 / Chapter 6.0.1 --- Conclusions --- p.94 / Chapter 6.0.2 --- Future Work --- p.96 / Bibliography --- p.97

With the booming development of WWW, many search engines have been developed to help users to find useful information from a great quantity of data. However, users may have different needs in different situations. Opposite to the Information Retrieval where users retrieve data actively, Information Filtering (IF) sends information from servers to passive users through broadcast mediums, rather than being searched by them. Therefore, each user has his (or her) profile stored in the database, where a profile records a set of interest items that can present his (or her) interests or habits. To efficiently store many user profiles in servers and filter irrelevant users, many signature-based index techniques are applied in IF systems. By using signatures, IF does not need to compare each item of profiles to filter out irrelevant ones. However, because signatures are incomplete information of profiles, it is very hard to answer the complex queries by using only the signatures. Therefore, a critical issue of the signature-based IF service is how to index the signatures of user profiles for an efficient filtering process. There are often two types of queries in the signature-based IF systems, the inexact filtering and the similarity search queries. In the inexact filtering, a query is an incoming document and it needs to find the profiles whose interest items are all included in the query. On the other hand, in the similarity search, a query is a user profile and it needs to find the users whose interest items are similar to the query user. In this thesis, we propose an ID-tree index strategy, which indexes signatures of user profiles by partitioning them into subgroups using a binary tree structure according to all of the different items among them. Basically, our ID-tree index strategy is a kind of the signature tree. In an ID-tree, each path from the root to a leaf node is the signature of the profile pointed by the leaf node. Because each profile is pointed only by one leaf node of the ID-tree, there will be no collision in the structure. In other words, there will be no two profiles assigned to the same signature. Moreover, only the different items among subgroups of profiles will be checked at one time to filter out irrelevant profiles for queries. Therefore, our strategy can answer the inexact filtering and the similarity search queries with less number of accessed profiles as compared to the previous strategies. Moreover, to build the index of signatures, it needs less time to batch a great deal of database profiles. From our simulation results, we show that our strategy can access less number of profiles to answer the queries than Chen's signature tree strategy for the inexact filtering and Aggarwal et al.'s SG-table strategy for the similarity search.

none

Owing to the booming development of the WWW, it creates many new challenges for information filtering. Information Filtering (IF) is an area of research that develops tools for discriminating between relevant and irrelevant information. IF can find good matches between the web pages and the users' information needs. Users first give descriptions about what they need, i.e., user profiles, to start the services. A profile index is built on these profiles. A series of incoming web pages will be put into the matching process. Each incoming web page is represented in the same form of the user profile. In this way, the users who are interested in an incoming web page can be identified by comparing the descriptions of the web page with each user profile. At last, the web page will be recommended to the users whose profiles belong to the filtered results. Therefore, a critical issue of the information filtering service is how to index the user profiles for an efficient matching process. When we index the user profile, we can reduce the costs of storage space and the processing time for modifying the user profiles. In this thesis, first, we propose a count-based tree method, which takes the count of each keyword into consideration, to reduce the large storage spaces as needed by the tree method. Next, three large-itemset-based methods are proposed to reduce the storage space, which are called the count-major large itemset method, the weighted large itemset method and the hybrid method. In these three large-itemset-based methods, we first cluster profiles with similar interests into the same group. Next, for each cluster, we apply the mining association rules techniques to help us to construct the index strategies. We design three methods by using the idea of the Apriori algorithm which is one of well-known approaches in mining association rules. But, we modify the minimum support and the goal in the Apriori algorithm. We may not always output the large itemset Lk. That is, we may only use Lw, where w < k. In summary, the cost of storage spaces of our four methods are less than that of the tree method proposed by Yan and Garcia-Molina. According to our simulation results, each of our four methods may provide the best result when different input data sets. Next, we propose a large-itemset-based approach to the incremental update of the index structure for storing keywords to reduce the update cost. When someone's interests are often changed, we must care about the way how to provide the low update cost of the index structure. We take the weight of each keyword into consideration. That is, each keyword can be distinguished the long-term interest which has weight above the threshold from the short-term interest which has weight below the threshold. Owing to that the probability of modifying the short-term interests is higher than that of modifying the long-term interests, we can update the short-term interests locally. According to our simulation results, our method really can reduce the update cost as needed by Wu and Chen' methods.