Xueqing Gong

Search Behavior Based Latent Semantic User Segmentation for Advertising Targeting

The popularity of internet usage greatly motivates the online advertising activities. Compared to advertising on traditional media, online advertising has rich information as well as necessary techniques to achieve precise user targeting. This rich information includes the search behaviors of a user, such as queries issued, or the ads clicked by the user. For popular websites with large number of active users, ad delivery targeting at individual users puts too much burden on the system. User segmentation is an alternative way to relieve this burden by grouping users of similar interests together, then the ad delivery system targets the user segments to display relevant ads, instead of individual users. Existing user segmentation work either adapts clustering methods without considering the hidden semantics embedded in the data, such as K-means, or treats users as data instance and clusters users indirectly even if the latent semantics is incorporated into the transformed data, such as PLSA or LDA. In this paper, we present a search behavior based latent semantic user segmentation method and validate its effectiveness on new ads. Instead of treating users as data instances, they are used as attributes of user issued queries or clicked ads which are considered to be data instances. LDA is then applied to this data set to directly obtain the user segments. Compared to popular K-means clustering, our approach achieves higher CTR values on new ads, with only simple search information.

Practical Duplicate Bug Reports Detection in a Large Web-Based Development Community

Most of large web-based development communities require a bug tracking system to keep track of various bug reports. However, duplicate bug reports tend to result in waste of resources, and may cause potential conflicts. There have been two types of works focusing on this problem: relevant bug report retrieval [8][11][10][13] and duplicate bug report identification [5][12]. The former methods can achieve high accuracy (82%) in the top 10 results in some dataset, but they do not really reduce the workload of developers. The latter methods still need further improvement on the performance.

Discovering collective viewpoints on micro-blogging events based on community and temporal aspects

Towards hot events, microblogs usually collect diverse and abundant thoughts, comments and opinions in a short period. It is interesting and meaningful to find how users are thinking about such events. In this paper, we aim to mine collective viewpoints from micro-blogging messages for any given event. Since a user can post multiple messages in a discussion, a user may have multiple viewpoints on a given event. Also user viewpoints may change under the influence of external events, such as news releases and activities, as time goes by. These present challenging of extracting collective viewpoints. To address this, we propose a T erm-T w eet- U ser (TWU ) graph, which simultaneously incorporates text content, community structure and temporal information, to model user postings over time. We first identify representative terms from tweets, which constitute collective viewpoints. And then we apply Random Walk on TWU graph to measure the relevance between terms and group them into collective viewpoints. Finally, we evaluated our approach based on 817,422 tweets collected from Sina microblog, which is the biggest microblog in China. Experiments on the real dataset show the effectiveness of our model and algorithms.

Multi-dimensional data density estimation in P2P networks

Estimating the global data distribution in Peer-to-Peer (P2P) networks is an important issue and has not yet been well addressed. It can benefit many P2P applications, such as load balancing analysis, query processing, data mining, and so on. In this paper, we propose a novel algorithm which is based on compact multi-dimensional histogram information to achieve high estimation accuracy with low estimation cost. Maintaining data distribution in a multi-dimensional histogram which is spread among peers without overlapping and each part of which is further condensed by a set of discrete cosine transform coefficients, each peer is capable to hierarchically accumulate the compact information to the entire histogram by information exchange and consequently estimates the global data density with accuracy and efficiency. Algorithms on discrete cosine transform coefficients hierarchically accumulating as well as density estimation error are introduced with detailed theoretical analysis and proof. Our extensive performance study confirms the effectiveness and efficiency of our methods on density estimation in dynamic P2P networks.

TaxiHailer: A Situation-Specific Taxi Pick-Up Points Recommendation System

This demonstration presents TaxiHailer, a situation-specific recommendation system for passengers who are eager to find a taxi. Given a query with departure point, destination and time, it recommends pick-up points within a specified distance and ranked by potential waiting time. Unlike existing works, we consider three sets of features to build regression models, as well as Poisson process models for road segment clusters. We evaluate and choose the most proper models for each cluster under different situations. Also, TaxiHailer gives destination-aware recommendations for pick-up points with driving directions. We evaluate our recommendation results based on real GPS datasets.

Real-Time Search over a Microblogging System

The microblog systems are becoming more and more popular in recent years, including twitter and sina weibo, etc. Users are not only used to release the latest news, but also willing to search interesting topics in the system. However, two main issues make real-time search over a microblog system challenging. First, the volume of data in a system is quite huge. Second, the users are eager to get the response in short time. In this paper, we have designed and implemented a real-time search system for a microblog system, which includes four key modules, including data preprocessing, indexing, searching and ranking. The extensive experiments on the real datasets demonstrate the effectiveness and efficiency of our proposed method.

Shortest Path Based Potential Common Friend Recommendation in Social Networks

Friend recommendation is a very popular application in online social network(OSN) services to help users make new friends and expand their social circle. Much of the existing research is based on either topological structure of OSN or user profiles to recommend new friends, while research based on both topological structure and user profiles to recommend potentially common friends between two undirectly connected users is still lacking. Recommending potentially common friends is useful in practical applications such as finding potential partners between two business groups in commercial social network in order to expand their business range or to reach potential customers, finding common bus stops between two specified bus stops on traffic network etc. In this paper we propose a novel shortest path based common friends recommendation approach in OSN to find and recommend potential common friends between two users. Our approach consists of the improved Floyd_Warshall algorithm and the Extended Longest Common Subsequence(ELCS) algorithm on both the topological structures of OSN and the partial profiles of users. The experiment results show that our approach can help users find the potential common friends efficiently and effectively.

Detecting anomaly in data streams by fractal model

Detecting anomaly in data streams attracts great attention in both academic and industry communities due to its wide range application in venture analysis, network monitoring, trend analysis and so on. However, existing methods on anomaly detection suffer three problems. 1) A large number of false positive results are generated. 2) Training data are needed to build the detection model, and an appropriate time window size along with corresponding threshold has to be set empirically. 3) Both time and space overhead is usually very high. To address these limitations. We propose a fractal-model-based approach to detection of anomalies that change underlying data distribution in this paper. Both a history-based algorithm and a parameter-free algorithm are introduced. We show that the later method consumes only limited memory and does not involve any training process. Theoretical analyses of the algorithm are presented. The experimental results on real life data sets indicate that, compared with existing anomaly detection methods, our algorithm can achieve higher precision with less space and time complexity.

ACID Encountering the CAP Theorem: Two Bank Case Studies

In the era of big data, we may adopt the distributed architecture for a transaction processing system due to some reasons, including distributed branches, heavy demand and operational expenditure, etc. In terms of the CAP Theorem, a transaction processing system associated with ACID properties is infeasible to work well in the distributed architecture. It is indispensable to address how to make a trade-off between availability and partition tolerance for a bank as it favors the consistency in the distributed system. In this research, we conduct two case studies to address the question using two transaction logs collected from a bank in China. We mainly analyze the table dependency and the table concurrency, and find that (1) it is arduous to partition the data in the database system associated with ACID properties, (2) in-memory architecture for updating transactions may be an alternative for building a transaction processing system.

Real-time and Personalized Search over a Microblogging System

As the microblogging systems such as Twitter and Sina Weibo become more and more popular in recent years, the requirement for real-time and personalized search over microblogging systems also becomes more important. In general, a user may expect a quick response that also satisfies her personalized requirements. Unfortunately, since there exist a huge number of users and massive updating microblogs in a microblogging system, personalized search on the system becomes a challenging task. In this paper, we design a new search engine containing four modules to infer the topics of microblogs and update the interests of users, build indexes efficiently, return microblogs for a keyword search, and personalize the order of microblogs respectively. We also conduct a series of experiments on a real dataset to illustrate the effectiveness and efficiency of the proposed methods.