Concept drift refers to a non stationary learning problem over time. The training and the application data often mismatch in real life problems. In this report we present a context of concept drift problem 1. We focus on the issues relevant to adaptive training set formation. We present the framework and terminology, and formulate a global picture of concept drift learners design. We start with formalizing the framework for the concept drifting data in Section 1. In Section 2 we discuss the adaptivity mechanisms of the concept drift learners. In Section 3 we overview the principle mechanisms of concept drift learners. In this chapter we give a general picture of the available algorithms and categorize them based on their properties. Section 5 discusses the related research fields and Section 5 groups and presents major concept drift applications. This report is intended to give a bird's view of concept drift research field, provide a context of the research and position it within broad spectrum of research fields and applications.

In the form of topic discussions, users interact with each other to share knowledge and exchange information in online forums. Modeling the evolution of topic discussion reveals how information propagates on Internet and can thus help understand sociological phenomena and improve the performance of applications such as recommendation systems. In this paper, we argue that a user’s participation in topic discussions is motivated by either her friends or her own preferences. Inspired by the theory of information flow, we propose dynamic topic discussion models by mining influential relationships between users and individual preferences. Reply relations of users are exploited to construct the fundamental influential social network. The property of discussed topics and time lapse factor are also considered in our modeling. Furthermore, we propose a novel measure called ParticipationRank to rank users according to how important they are in the social network and to what extent they prefer to participate in the discussion of a certain topic. The experiments show our model can simulate the evolution of topic discussions well and predict the tendency of user’s participation accurately.

With the increasing popularity of location tracking services such as GPS, more and more mobile data are being accumulated. Based on such data, a potentially useful service is to make timely and targeted recommendations for users on places where they might be interested to go and activities that they are likely to conduct. For example, a user arriving in Beijing might wonder where to visit and what she can do around the Forbidden City. A key challenge for such recommendation problems is that the data we have on each individual user might be very limited, while to make useful and accurate recommendations, we need extensive annotated location and activity information from user trace data. In this paper, we present a new approach, known as user-centered collaborative location and activity filtering (UCLAF), to pull many users’ data together and apply collaborative filtering to find like-minded users and like-patterned activities at different locations. We model the userlocation- activity relations with a tensor representation, and propose a regularized tensor and matrix decomposition solution which can better address the sparse data problem in mobile information retrieval. We empirically evaluate UCLAF using a real-world GPS dataset collected from 164 users over 2.5 years, and showed that our system can outperform several state-of-the-art solutions to the problem.

Matrix factorization is a fundamental technique in machine learning that is applicable to collaborative filtering, information retrieval and many other areas. In collaborative filtering and many other tasks, the objective is to fill in missing elements of a sparse data matrix. One of the biggest challenges in this case is filling in a column or row of the matrix with very few observations. In this paper we introduce a Bayesian matrix factorization model that performs regression against side information known about the data in addition to the observations. The side information helps by adding observed entries to the factored matrices. We also introduce a nonparametric mixture model for the prior of the rows and columns of the factored matrices that gives a different regularization for each latent class. Besides providing a richer prior, the posterior distribution of mixture assignments reveals the latent classes. Using Gibbs sampling for inference, we apply our model to the Netflix Prize problem of predicting movie ratings given an incomplete user-movie ratings matrix. Incorporating rating information with gathered metadata information, our Bayesian approach outperforms other matrix factorization techniques even when using fewer dimensions.

Data sparsity is a major problem for collaborative filtering (CF) techniques in recommender systems, especially for new users and items. We observe that, while our target data are sparse for CF systems, related and relatively dense auxiliary data may already exist in some other more mature application domains. In this paper, we address the data sparsity problem in a target domain by transferring knowledge about both users and items from auxiliary data sources. We observe that in different domains the user feedbacks are often heterogeneous such as ratings vs. clicks. Our solution is to integrate both user and item knowledge in auxiliary data sources through a principled matrix-based transfer learning framework that takes into account the data heterogeneity. In particular, we discover the principle coordinates of both users and items in the auxiliary data matrices, and transfer them to the target domain in order to reduce the effect of data sparsity. We describe our method, which is known as coordinate system transfer or CST, and demonstrate its effectiveness in alleviating the data sparsity problem in collaborative filtering. We show that our proposed method can significantly outperform several state-of-the-art solutions for this problem.

Social tagging systems have emerged as an effective way for users to annotate and share objects on the Web. However, with the growth of social tagging systems, users are easily overwhelmed by the large amount of data and it is very difficult for users to dig out information that he/she is interested in. Though the tagging system has provided interest-based social network features to enable the user to keep track of other users' tagging activities, there is still no automatic and effective way for the user to discover other users with common interests. In this paper, we propose a User Recommendation (UserRec) framework for user interest modeling and interest-based user recommendation, aiming to boost information sharing among users with similar interests. Our work brings three major contributions to the research community: (1) we propose a tag-graph based community detection method to model the users' personal interests, which are further represented by discrete topic distributions; (2) the similarity values between users' topic distributions are measured by Kullback-Leibler divergence (KL-divergence), and the similarity values are further used to perform interest-based user recommendation; and (3) by analyzing users' roles in a tagging system, we find users' roles in a tagging system are similar to Web pages in the Internet. Experiments on tagging dataset of Web pages (Yahoo!~Delicious) show that UserRec outperforms other state-of-the-art recommender system approaches.

Collaborative filtering systems have been developed to manage information overload and improve discourse in online communities. In such systems, users rank content provided by other users on the validity or usefulness within their particular context. The goal is that "good" content will rise to prominence and "bad" content will fade into obscurity. These filtering mechanisms are not well-understood and have known weaknesses. For example, they depend on the presence of a large crowd to rate content, but such a crowd may not be present. Additionally, the community's decisions determine which voices will reach a large audience and which will be silenced, but it is not known if these decisions represent "the wisdom of crowds" or a "censoring mob." Our approach uses statistical machine learning to predict community ratings. By extracting features that replicate the community's verdict, we can better understand collaborative filtering, improve the way the community uses the ratings of their members, and design agents that augment community decision-making. Slashdot is an example of such a community where peers will rate each others' comments based on their relevance to the post. This work extracts a wide variety of features from the Slashdot metadata and posts' linguistic contents to identify features that can predict the community rating. We find that author reputation, use of pronouns, and author sentiment are salient. We achieve 76% accuracy predicting community ratings as good, neutral, or bad.

This research explored the transition of romantic relationships from meeting online to the first face-to-face date. It is inevitable that impressions of a partner will change to some degree, but how much, and with what consequences? One hundred and fifty users of a popular online dating site participated in the study. They recalled a person whom they had met through the site, reporting their impressions of their partners from both before and after the first face-to-face meeting. We expected, based on prior research demonstrating the importance of physical attractiveness in romantic attraction both on- and offline, that changes in beliefs about partners’ physical appeal would be the most powerful predictor of relationship longevity. However, they were unrelated to relationship success. Across all the dimensions we examined, impressions were in fact relatively stable, but when respondents said they knew their partners better after meeting face-to-face, relationships lasted longer.v

Given a set of reviews of products or merchants from a wide range of authors and several reviews websites, how can we measure the true quality of the product or merchant?  How do we remove the bias of individual authors or sources?  How do we compare reviews obtained from different websites, where ratings may be on different scales (1-5 stars, A/B/C, etc.)?  How do we filter out unreliable reviews to use only the ones with ``star quality''?  Taking into account these considerations, we analyze data sets from a variety of different reviews sites (the first paper, to our knowledge, to do this). These data sets include 8 million product reviews and 1.5 million merchant reviews. We explore statistic- and heuristic- based models for estimating the true quality of a product or merchant, and compare the performance of these estimators on the task of ranking pairs of objects.  We also apply the same models to the task of using Netflix ratings data to rank pairs of movies, and discover that the performance of the different models is surprisingly similar on this data set.

We propose a new method of recommending questions to answerers so as to suit the answerers’ knowledge and interests in User-Interactive Question Answering (QA) communities. A question recommender can help answerers select the questions that interest them. This increases the number of answers, which will activate QA communities. An effective question recommender should satisfy the following three requirements: First, its accuracy should be higher than the existing category-based approach; more than 50% of answerers select the questions to answer according a fixed system of categories. Second, it should be able to recommend unanswered questions because more than 2,000 questions are posted every day. Third, it should be able to support even those people who have never answered a question previously, because more than 50% of users in current QA communities have never given any answer. To achieve an effective question recommender, we use question histories as well as the answer histories of each user by combining collaborative filtering schemes and content-base filtering schemes. Experiments on real log data sets of a famous Japanese QA community, Oshiete goo, show that our recommender satisfies the three requirements.