Changping Wang

MapDupReducer: detecting near duplicates over massive datasets

Near duplicate detection benefits many applications, e.g., on-line news selection over the Web by keyword search. The purpose of this demo is to show the design and implementation of MapDupReducer, a MapReduce based system capable of detecting near duplicates over massive datasets efficiently.

Inferring continuous dynamic social influence and personal preference for temporal behavior prediction

It is always attractive and challenging to explore the intricate behavior data and uncover peoples motivations, preference and habits, which can greatly benefit many tasks including link prediction, item recommendation, etc. Traditional work usually studies peoples behaviors without time information in a static or discrete manner, assuming the underlying factors stay invariant in a long period. However, we believe peoples behaviors are dynamic, and the contributing factors including the social influence and personal preference for behaviors are varying continuously over time. Such continuous dynamics convey important knowledge about peoples behavior patterns; ignoring them would lead to inaccurate models. In this work, we address the continuous dynamic modeling of temporal behaviors. To model the fully continuous temporal dynamics of behaviors and the underlying factors, we propose the DP-Space, a dynamic preference probability space, which can capture their smooth variation in various shapes over time with flexible basis functions. Upon that we propose a generative dynamic behavior model, ConTyor, which considers the temporal item-adoption behaviors as joint effect of dynamic social influence and varying personal preference over continuous time. We also develop effective inference methods for ConTyor and present its applications. We conduct a comprehensive experimental study using real-world datasets to evaluate the effectiveness of our model and the temporal modeling. Results verify that ConTyor outperforms existing state-of-the-art static and temporal models in behavior predictions. Moreover, in our detailed study on temporal modeling, we show that temporal modeling is superior to static approaches and modeling over continuous time is further better than that over discrete time. We also demonstrate that the ancient behavior data can still become important and beneficial if modeled well.

Inferring Directions of Undirected Social Ties

The directionality is a significant but inherent property of social ties, though usually ignored in undirected social networks due to its invisibility. However, we believe most social ties are natively directed, and the perception of directionality can improve our understanding about the network structures and further benefit other tasks upon social networks. In this study, we address the latent tie direction inference problem in undirected social networks. We engage in the investigation of directionality on real-world large-scale directed social networks and summarize our findings using four patterns. Upon that we propose a family of ReDirect approaches, including ReDirect-N, ReDirect-T and ReDirect-One, to inferring the hidden directions of undirected social ties based on the network topology only. ReDirect can incorporate with other predictive tasks, and introduce supervision to improve performance. We also present a simple but effective strategy to construct self-labeled data. Experimental results show that even without external information, our approach can recover the directions of networks effectively. Moreover, we find the ReDirect approaches can benefit the predictive tasks remarkably in an experimental study on link prediction. The ReDirect family can be a beneficial general data preprocess tool for various network analysis tasks by uncovering the hidden directions.

An attribute-based community search method with graph refining

In many complex networks, there exist diverse network topologies as well as node attributes. However, the state-of-the-art community search methods which aim to find out communities containing the query nodes only consider the network topology, but ignore the effect of node attributes. This may lead to the inaccuracy of the predicted communities. In this paper, we propose an attribute-based community search method with graph refining technique, called AGAR. First, we present the concepts of topology-based similarity and attribute-based similarity to construct a TA-graph. The TA-graph can reflect both the relations between nodes from the respect of the network topology and that of the node attributes. Then, we construct AttrTCP-index based on the structure of TA-graph. Finally, by querying the AttrTCP-index, we can find out the communities for the query nodes. Experimental results on real-world networks demonstrate AGAR is an effective and efficient community search method by considering both the network topology and node attributes.

Preference Join on Heterogeneous Data

There are different types of join operation dealing with different issues in database research. However, existing join operations cannot meet the increasing demands of the real world. In this paper, we define a new join operation, the preference join (p-join), which introduces the concepts of the personal preference and the satisfaction operator on various data types. We present a general join algorithm (Nested Loop) to deal with the p-join, and we also propose an advanced algorithm called MFV for p-join. To improve the MFV algorithm, two enhanced mapping methods are employed. A large number of experiments on both real-world and synthetic data sets are conducted. The experimental results demonstrate the effectiveness, efficiency and scalability of our methods, and show the advanced algorithms have advantages over the general algorithms.

AGAR: an attribute-based graph refining method for community search

In many complex networks, there exist diverse network topologies as well as node attributes. However, the state-of-the-art community search methods which aim to find out communities containing query nodes, only consider the network topology but ignore the effect of node attributes. This may lead to the inaccuracy of communities, especially in sparse networks. In this paper, we propose an attribute-based graph refining method called AGAR for community search. Our method refines the graph topology based on both graph topologies and node attributes, and then helps community search methods obtain more accurate and meaningful communities. Experimental results on two real-world networks demonstrate AGAR can refine the initial graph into a more meaningful graph and help community search methods to find more accurate communities.

Efficient Computation of G-Skyline Groups

The skyline of a data point set is made up of the best points in the set, and is very important for multi-criteria decision making. In these years, the skyline problem attracts more and more attention, and many variants of the traditional skyline emerge in the database field. One recent and important variant is group-based skyline, which aims to find the best groups of points in a given set. In this paper, we bring forward an efficient approach, called minimum dominance search (MDS), to solve the g-skyline problem, a latest group-based skyline problem. MDS consists of two steps: In the first step, we construct a novel g-skyline support structure, i.e., minimum dominance graph (MDG), which proves to be a minimum g-skyline support structure. In the second step, we search for g-skyline groups based on the MDG through two searching algorithms, and a skyline-combination based optimization strategy is employed to improve these two algorithms. We conduct comprehensive experiments on both synthetic and real-world data sets, and show that our algorithms are orders of magnitude faster than the state-of-the-art in most cases.

MAVis: A Multiple Microblogs Analysis and Visualization Tool

An increasing number of people obtain and share information on social networks through short text messages, a.k.a. microblogs. These microblogs propagate widely online based on the followship between users as well as the retweeting mechanism. The regular pattern of retweeting behaviors can be discovered by mining the historical retweet data, and the key users in the information diffusion process can also be found in this way. This paper gives the novel definition of information diffusion network and three categories of nodes in the network. A tool designed to mine the information diffusion network and visualize the result is implemented. This paper introduces related definitions, the architecture, mining algorithms and the visualization interface.