UTS site search

Dr Ling Chen

Biography

Dr Ling Chen is a Senior Lecturer with the UTS Priority Research Centre for Artificial Intelligence (CAI) and the Faculty of Engineering and Information Technology (FEIT) in the University of Technology Sydney (UTS). She is currently the director of the Data Science & Knowledge Discovery Laboratory (The DSKD Lab). Before joining UTS, she was a Postdoc Research Fellow with L3S Research Center, Leibiniz University Hannover, Germany. Dr Chen received her PhD in 2008 in Computer Engineering, from Nanyang Technological University Singapore.

Dr Chen's main research interests focus on data mining and machine learning. She has worked on fundamental data mining tasks such as pattern mining from structured data or uncertain data and hash function design. Her recent work also includes knowledge discovery from social networks and social media. She has developed novel and effective algorithms for event detection from social media and recommendation in social networks.

Dr Chen has published more than 60 papers in major international data mining conferences including SIGKDD, SIGMOD, VLDB, ICDM, SDM and CIKM, and journals like ACM TOIS and IEEE TKDE. Dr Chen has served as a PC member for many conferences such as SIGKDD, AAAI and IJCAI. She was the Conference Chair of AusDM 2015.

Senior Lecturer, A/DRsch Centre for Artificial Intelligence
Core Member, AAI - Advanced Analytics Institute
Core Member, Centre for Artificial Intelligence
Philosophy
 
Phone
+61 2 9514 1925

Research Interests

  • Data mining and machine learning
  • Social network analysis and social media mining
  • Information retrieval and recommendation
Can supervise: Yes
Registered at Level 1.

Conferences

Wu, W., Li, B., Chen, L. & Zhang, C. 2016, 'Cross-View Feature Hashing for Image Retrieval', Advances in Knowledge Discovery and Data Mining, Pacific Asia Knowledge Discovery and Data Mining Conference (PAKDD) 2016, Springer International Publishing, The University of Auckland Auckland, New Zealand, pp. 203-214.
View/Download from: Publisher's site
Traditional cross-view information retrieval mainly rests on correlating two sets of features in different views. However, features in different views usually have different physical interpretations. It may be inappropriate to map multiple views of data onto a shared feature space and directly compare them. In this paper, we propose a simple yet effective Cross-View Feature Hashing (CVFH) algorithm via a 'partition and match approach. The feature space for each view is bi-partitioned multiple times using B hash functions and the resulting binary codes for all the views can thus be represented in a compatible B-bit Hamming space. To ensure that hashed feature space is effective for supporting generic machine learning and information retrieval functionalities, the hash functions are learned to satisfy two criteria: (1) the neighbors in the original feature spaces should be also close in the Hamming space; and (2) the binary codes for multiple views of the same sample should be similar in the shared Hamming space. We apply CVFH to crossview image retrieval. The experimental results show that CVFH can outperform the Canonical Component Analysis (CCA) based cross-view method.
Wu, W., Li, B., Chen, L. & Zhang, C. 2016, 'Cross-view feature hashing for image retrieval', Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), Pacific-Asia Conference on Knowledge Discovery and Data Mining (PAKDD), Auckland, New Zealand, pp. 203-214.
View/Download from: UTS OPUS or Publisher's site
© Springer International Publishing Switzerland 2016.Traditional cross-view information retrieval mainly rests on correlating two sets of features in different views. However, features in different views usually have different physical interpretations. It may be inappropriate to map multiple views of data onto a shared feature space and directly compare them. In this paper, we propose a simple yet effective Cross-View Feature Hashing (CVFH) algorithm via a 'partition and match approach. The feature space for each view is bi-partitioned multiple times using B hash functions and the resulting binary codes for all the views can thus be represented in a compatible B-bit Hamming space. To ensure that hashed feature space is effective for supporting generic machine learning and information retrieval functionalities, the hash functions are learned to satisfy two criteria: (1) the neighbors in the original feature spaces should be also close in the Hamming space; and (2) the binary codes for multiple views of the same sample should be similar in the shared Hamming space. We apply CVFH to cross view image retrieval. The experimental results show that CVFH can outperform the Canonical Component Analysis (CCA) based cross-view method.
Pang, G., Cao, L. & Chen, L. 2016, 'Outlier detection in complex categorical data by modelling the feature value couplings', Proceedings of the 25th International Joint Conference on Artificial Intelligence, AAAI, pp. 1902-1908.
Han, B., Tsang, I.W. & Chen, L. 2016, 'On the Convergence of A Family of Robust Losses for Stochastic Gradient Descent', Machine Learning and Knowledge Discovery in Databases - LNCS, The European Conference on Machine Learning and Principles and Practice of Knowledge Discovery (ECML-PKDD), Springer, Riva del Garda, Italy.
View/Download from: UTS OPUS or Publisher's site
The convergence of Stochastic Gradient Descent (SGD) using convex loss functions has been widely studied. However, vanilla SGD methods using convex losses cannot perform well with noisy labels, which adversely affect the update of the primal variable in SGD methods. Unfortunately, noisy labels are ubiquitous in real world applications such as crowdsourcing. To handle noisy labels, in this paper, we present a family of robust losses for SGD methods. By employing our robust losses, SGD methods successfully reduce negative effects caused by noisy labels on each update of the primal variable. We not only reveal the convergence rate of SGD methods using robust losses, but also provide the robustness analysis on two representative robust losses. Comprehensive experimental results on six real-world datasets show that SGD methods using robust losses are obviously more robust than other baseline methods in most situations with fast convergence.
Liu, B., Chen, L., Liu, C., Zhang, C. & Qiu, W. 2016, 'Mining co-locations from continuously distributed uncertain spatial data', Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics):Web Technologies and Applications 18th Asia-Pacific Web Conference, APWeb 2016, Web Technologies and Applications : Asia-Pacific Web Conference (APWeb), Springer, Suzhou, China, pp. 66-78.
View/Download from: UTS OPUS or Publisher's site
© Springer International Publishing Switzerland 2016.A co-location pattern is a group of spatial features whose instances tend to locate together in geographic space. While traditional co-location mining focuses on discovering co-location patterns from deterministic spatial data sets, in this paper, we study the problem in the context of continuously distributed uncertain data. In particular, we aim to discover co-location patterns from uncertain spatial data where locations of spatial instances are represented as multivariate Gaussian distributions. We first formulate the problem of probabilistic co-location mining based on newly defined prevalence measures. When the locations of instances are represented as continuous variables, the major challenges of probabilistic co-location mining lie in the efficient computation of prevalence measures and the verification of the probabilistic neighborhood relationship between instances. We develop an effective probabilistic co-location mining framework integrated with optimization strategies to address the challenges. Our experiments on multiple datasets demonstrate the effectiveness of the proposed algorithm.
Wang, W., Yin, H., Sadiq, S., Chen, L., Xie, M. & Zhou, X. 2016, 'SPORE: A Sequential Personalized Spatial Item Recommender System', IEEE International Conference on Data Engineering (ICDE), Helsinki.
Song, K., Chen, L., Gao, W., Feng, S., Wang, D. & Zhang, C. 2016, 'PerSentiment: A Personalized Sentiment Classi cation System for Microblog Users', International World Wide Web Conference (WWW), Montreal.
Liu, C. & Chen, L. 2016, 'Summarizing Uncertain Transaction Databases by Probabilistic Tiles', The annual International Joint Conference on Neural Networks (IJCNN), Vancouver.
Song, K., Gao, W., Chen, L., Feng, S., Wang, D. & Zhang, C. 2016, 'Build Emotion Lexicon from the Mood of Crowd via Topic-Assisted Joint Non-negative Matrix Factorization', ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR), Pisa, Italy.
Pang, G., Cao, L. & Chen, L. 2016, 'Identifying Outliers in Complex Categorical Data by Modeling the Feature Value Couplings', International Joint Conference on Artificial Intelligence (IJCAI).
Pang, G., Cao, L., Chen, L. & Liu, H. 2016, 'Unsupervised Feature Selection for Outlier De- tection by Modelling Hierarchical Value-Feature Couplings', The IEEE International Conference on Data Mining (ICDM), Barcelona.
View/Download from: UTS OPUS
Wu, W., Li, B., Chen, L. & Zhang, C. 2016, 'Canonical Consistent Weighted Sampling for Real-Value Weighted Min-Hash', Proceedings of the 2016 IEEE 16th International Conference on Data Mining, The IEEE International Conference on Data Mining (ICDM), IEEE, Barcelona, Spain, pp. 1287-1292.
View/Download from: UTS OPUS or Publisher's site
Min-Hash, as a member of the Locality Sensitive Hashing (LSH) family for sketching sets, plays an important role in the big data era. It is widely used for efficiently estimating similarities of bag-of-words represented data and has been extended to dealing with multi-sets and real-value weighted sets. Improved Consistent Weighted Sampling (ICWS) has been recognized as the state-of-the-art for real-value weighted Min-Hash. However, the algorithmic implementation of ICWS is flawed because it violates the uniformity of the Min-Hash scheme. In this paper, we propose a Canonical Consistent Weighted Sampling (CCWS) algorithm, which not only retains the same theoretical complexity as ICWS but also strictly complies with the definition of Min-Hash. The experimental results demonstrate that the proposed CCWS algorithm runs faster than the state-of-the-arts while achieving similar classification performance on a number of real-world text data sets.
Liu, B., Chen, L., Liu, C., Zhang, C. & Qiu, W. 2015, 'RCP Mining: Towards the Summarization of Spatial Co-location Patterns', Advances in Spatial and Temporal Databases (LNCS), 14th International Symposium on Advances in Spatial and Temporal Databases, Springer, Hong Kong, China, pp. 451-469.
View/Download from: UTS OPUS or Publisher's site
Co-location pattern mining is an important task in spatial data mining. However, the traditional framework of co-location pattern mining produces an exponential number of patterns because of the downward closure property, which makes it hard for users to understand, or apply. To address this issue, in this paper, we study the problem of mining representative co-location patterns (RCP). We first define a covering relationship between two co-location patterns by finding a new measure to appropriately quantify the distance between patterns in terms of their prevalence, based on which the problem of RCP mining is formally formulated. To solve the problem of RCP mining, we first propose an algorithm called RCPFast, adopting the post-mining framework that is commonly used by existing distance-based pattern summarization techniques. To address the peculiar challenge in spatial data mining, we further propose another algorithm, RCPMS, which employs the mine-and-summarize framework that pushes pattern summarization into the co-location mining process. Optimization strategies are also designed to further improve the performance of RCPMS. Our experimental results on both synthetic and real-world data sets demonstrate that RCP mining effectively summarizes spatial co-location patterns, and RCPMS is more efficient than RCPFast, especially on dense data sets.
Wang, H., Zhang, P., Chen, L., Liu, H. & Zhang, C. 2015, 'Online Diffusion Source Detection in Social Networks', Proceedings of the 2015 International Joint Conference on Neural Networks (IJCNN), International Joint Conference on Neural Networks (IJCNN), IEEE, Killarney, Ireland, pp. 1-8.
View/Download from: Publisher's site
In this paper we study a new problem of online diffusion source detection in social networks. Existing work on diffusion source detection focuses on offline learning, which assumes data collected from network detectors are static and a snapshot of network is available before learning. However, an offline learning model does not meet the needs of early warning, real-time awareness, and real-time response of malicious information spreading in social networks. In this paper, we combine online learning and regression-based detection methods for real-time diffusion source detection. Specifically, we propose a new 1 non-convex regression model as the learning function, and an Online Stochastic Sub-gradient algorithm (OSS for short). The proposed model is empirically evaluated on both synthetic and real-world networks. Experimental results demonstrate the effectiveness of the proposed model.
Wang, W., Yin, H., Chen, L., Sun, Y., Sadiq, S. & Zhou, X. 2015, 'GEO-SAGE: A geographical sparse additive generative model for spatial item recommendation', The 21th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, Sydney, Australia, pp. 1255-1264.
View/Download from: UTS OPUS
Song, K., Feng, S., Gao, W., Wang, D., Chen, L. & Zhang, C. 2015, 'Building emotion lexicon from microblogs by combining effects of seed words and emoticons in a heterogeneous graph', Proceedings of the 26th ACM Conference on Hypertext & Social Media, The 26th ACM Conference on HyperText and Social Media (HT'15), ACM, Guzelyurt, Northern Cyprus, pp. 283-292.
View/Download from: Publisher's site
As an indispensable resource for emotion analysis, emotion lexicons have attracted increasing attention in recent years. Most existing methods focus on capturing the single emotional effect of words rather than the emotion distributions which are helpful to model multiple complex emotions in a subjective text. Meanwhile, automatic lexicon building methods are overly dependent on seed words but neglect the effect of emoticons which are natural graphical labels of fine-grained emotion. In this paper, we propose a novel emotion lexicon building framework that leverages both seed words and emoticons simultaneously to capture emotion distributions of candidate words more accurately. Our method overcomes the weakness of existing methods by combining the effects of both seed words and emoticons in a unified three-layer heterogeneous graph, in which a multi-label random walk (MLRW) algorithm is performed to strengthen the emotion distribution estimation. Experimental results on real-world data reveal that our constructed emotion lexicon achieves promising results for emotion classification compared to the state-of-the-art lexicons
Wang, H., Zhang, P., Tsang, I., Chen, L. & Zhang, C. 2015, 'Defragging Subgraph Features for Graph Classification', Proceedings of the 24th ACM International on Conference on Information and Knowledge Management, The 24th ACM International on Conference on Information and Knowledge Management (CIKM'15), ACM, Melbourne, VIC, Australia, pp. 1687-1690.
View/Download from: UTS OPUS or Publisher's site
Graph classification is an important tool for analysing structured and semi-structured data, where subgraphs are commonly used as the feature representation. However, the number and size of subgraph features crucially depend on the threshold parameters of frequent subgraph mining algorithms. Any improper setting of the parameters will generate many trivial short-pattern subgraph fragments which dominate the feature space, distort graph classifiers and bury interesting long-pattern subgraphs. In this paper, we propose a new Subgraph Join Feature Selection (SJFS) algorithm. The SJFS algorithm, by forcing graph classifiers to join short-pattern subgraph fragments, can defrag trivial subgraph features and deliver long-pattern interesting subgraphs. Experimental results on both synthetic and real-world social network graph data demonstrate the performance of the proposed method.
Wang, H., Zhang, P., Chen, L. & Zhang, C. 2015, 'Socialanalysis: A Real-Time query and mining system from social media data streams', Databases Theory and Applications (LNCS), Australasian Database Conference, Springer, Melbourne, Australia, pp. 318-322.
View/Download from: Publisher's site
© Springer International Publishing Switzerland 2015. In this paper, we present our recent progress of designing a real-time system, SocialAnalysis, to discover and summarize emergent social events from social media data streams. In social networks era, people always frequently post messages or comments about their activities and opinions. Hence, there exist temporal correlations between the physical world and virtual social networks, which can help us to monitor and track social events, detecting and positioning anomalous events before their outbreakings, so as to provide early warning. The key technologies in the system include: (1) Data denoising methods based on multi-features, which screens out the query-related event data from massive background data. (2) Abnormal events detection methods based on statistical learning, which can detect anomalies by analyzing and mining a series of observations and statistics on the time axis. (3) Geographical position recognition, which is used to recognize regions where abnormal events may happen.
Yin, H., Cui, B., Chen, L., Hu, Z. & Huang, Z. 2014, 'A temporal context-aware model for user behavior modeling in social media systems', International Conference on Management of Data, SIGMOD 2014, International Conference on Management of Data, SIGMOD 2014,, ACM, Snowbird, UT, USA, pp. 1543-1554.
View/Download from: UTS OPUS or Publisher's site
Wan, L., Chen, L. & Zhang, C. 2013, 'Mining Frequent Serial Episodes over Uncertain Sequence Data', The 16th International Conference on Extending Database Technology (EDBT 2013), International Conference on Extending Database Technology, ACM EDBT/ICDT 2013, Genoa, Italy, pp. 215-226.
View/Download from: UTS OPUS or Publisher's site
Data uncertainty has posed many unique challenges to nearly all types of data mining tasks, creating a need for uncertain data mining. In this paper, we focus on the particular task of mining probabilistic frequent serial episodes (P-FSEs) from uncertain sequence data, which applies to many real applications including sensor readings as well as customer purchase sequences. We first define the notion of P-FSEs, based on the frequentness probabilities of serial episodes under possible world semantics. To discover P-FSEs over an uncertain sequence, we propose: 1) an exact approach that computes the accurate frequentness probabilities of episodes; 2) an approximate approach that approximates the frequency of episodes using probability models; 3) an optimized approach that efficiently prunes a candidate episode by estimating an upper bound of its frequentness probability using approximation techniques. We conduct extensive experiments to evaluate the performance of the developed data mining algorithms. Our experimental results show that: 1) while existing research demonstrates that approximate approaches are orders of magnitudes faster than exact approaches, for P-FSE mining, the efficiency improvement of the approximate approach over the exact approach is marginal; 2) although it has been recognized that the normal distribution based approximation approach is fairly accurate when the data set is large enough, for P-FSE mining, the binomial distribution based approximation achieves higher accuracy when the the number of episode occurrences is limited; 3) the optimized approach clearly outperforms the other two approaches in terms of the runtime, and achieves very high accuracy.
Yin, H., Sun, Y., Cui, B., Hu, Z. & Chen, L. 2013, 'LCARS: A Location-Content-Aware Recommender System', ACM SIGKDD Conference on Knowledge Discovery and Data Mining, ACM International Conference on Knowledge Discovery and Data Mining, ACM, Chicago, Illinois USA, pp. 221-229.
View/Download from: UTS OPUS or Publisher's site
Newly emerging location-based and event-based social network services provide us with a new platform to understand users preferences based on their activity history. A user can only visit a limited number of venues/events and most of them are within a limited distance range, so the user-item matrix is very sparse, which creates a big challenge for traditional collaborative filtering-based recommender systems. The problem becomes more challenging when people travel to a new city where they have no activity history. In this paper, we propose LCARS, a location-content-aware recommender system that offers a particular user a set of venues (e.g., restaurants) or events (e.g., concerts and exhibitions) by giving consideration to both personal interest and local preference. This recommender system can facilitate peoples travel not only near the area in which they live, but also in a city that is new to them. Specifically, LCARS consists of two components: offline modeling and online recommendation. The offline modeling part, called LCA- LDA, is designed to learn the interest of each individual user and the local preference of each individual city by capturing item co- occurrence patterns and exploiting item contents. The online recommendation part automatically combines the learnt interest of the querying user and the local preference of the querying city to produce the top-k recommendations. To speed up this online process, a scalable query processing technique is developed by extending the classic Threshold Algorithm (TA). We evaluate the performance of our recommender system on two large-scale real data sets, Douban- Event and Foursquare. The results show the superiority of LCARS in recommending spatial items for users, especially when traveling to new cities, in terms of both effectiveness and efficiency.
Wan, L., Chen, L. & Zhang, C. 2013, 'Mining Dependent Frequent Serial Episodes from Uncertain Sequence Data', Proceedings of the13th IEEE International Conference on Data Mining, IEEE International Conference on Data Mining, IEEE Computer Society Press, Dallas, TX, USA, pp. 1211-1216.
View/Download from: UTS OPUS or Publisher's site
In this paper, we focus on the problem of mining Probabilistic Dependent Frequent Serial Episodes (P-DFSEs) from uncertain sequence data. By observing that the frequentness probability of an episode in an uncertain sequence is a Markov Chain imbeddable variable, we first propose an Embedded Markov Chain-based algorithm that efficiently computes the frequentness probability of an episode by projecting the probability space into a set of limited partitions. To further improve the computation efficiency, we devise an optimized approach that prunes candidate episodes early by estimating the upper bound of their frequentness probabilities.
Liu, C., Chen, L. & Zhang, C. 2013, 'Mining Probabilistic Representative Frequent Patterns From Uncertain Data', The 13th SIAM International Conference on Data Mining (SDM 2013), SIAM International Conference on Data Mining, SIAM / Omnipress, Austin, Texas, USA, pp. 1-9.
View/Download from: UTS OPUS or Publisher's site
Probabilistic frequent pattern mining over uncertain data has received a great deal of attention recently due to the wide applications of uncertain data. Similar to its counterpart in deterministic databases, however, probabilistic frequent pattern mining suffers from the same problem of generating an exponential number of result patterns. The large number of discovered patterns hinders further evaluation and analysis, and calls for the need to find a small number of representative patterns to approximate all other patterns. This paper formally defines the problem of probabilistic representative frequent pattern (P-RFP) mining, which aims to find the minimal set of patterns with sufficiently high probability to represent all other patterns. The problem's bottleneck turns out to be checking whether a pattern can probabilistically represent another, which involves the computation of a joint probability of supports of two patterns. To address the problem, we propose a novel and efficient dynamic programming-based approach. Moreover, we have devised a set of effective optimization strategies to further improve the computation efficiency. Our experimental results demonstrate that the proposed P-RFP mining effectively reduces the size of probabilistic frequent patterns. Our proposed approach not only discovers the set of P-RFPs efficiently, but also restores the frequency probability information of patterns with an error guarantee.
Liu, C., Chen, L. & Zhang, C. 2013, 'Summarizing Probabilistic Frequent Patterns: A Fast Approach', Proceedings of the 19th ACM SIGKDD international conference on Knowledge discovery and data mining (KDD13), ACM SIGKDD international conference on Knowledge discovery and data mining, ACM, Chicago, Illinois USA, pp. 527-535.
View/Download from: UTS OPUS or Publisher's site
Mining probabilistic frequent patterns from uncertain data has received a great deal of attention in recent years due to the wide applications. However, probabilistic frequent pattern mining suffers from the problem that an exponential number of result patterns are generated, which seriously hinders further evaluation and analysis. In this paper, we focus on the problem of mining probabilistic representative frequent patterns (P-RFP), which is the minimal set of patterns with adequately high probability to represent all frequent patterns. Observing the bottleneck in checking whether a pattern can probabilistically represent another, which involves the computation of a joint probability of the supports of two patterns, we introduce a novel approximation of the joint probability with both theoretical and empirical proofs. Based on the approximation, we propose an Approximate P-RFP Mining (APM) algorithm, which effectively and efficiently compresses the set of probabilistic frequent patterns. To our knowledge, this is the first attempt to analyze the relationship between two probabilistic frequent patterns through an approximate approach. Our experiments on both synthetic and real-world datasets demonstrate that the APM algorithm accelerates P-RFP mining dramatically, orders of magnitudes faster than an exact solution. Moreover, the error rate of APM is guaranteed to be very small when the database contains hundreds transactions, which further affirms APM is a practical solution for summarizing probabilistic frequent patterns.
Liu, C., Chen, L. & Zhang, C. 2013, 'Mining probabilistic representative frequent patterns from uncertain data', Proceedings of the 2013 SIAM International Conference on Data Mining, SDM 2013, pp. 73-81.
Copyright © SIAM. Probabilistic frequent pattern mining over uncertain data has received a great deal of attention recently due to the wide applications of uncertain data. Similar to its counterpart in deterministic databases, however, probabilistic frequent pattern mining suffers from the same problem of generating an exponential number of result patterns. The large number of discovered patterns hinders further evaluation and analysis, and calls for the need to find a small number of representative patterns to approximate all other patterns. This paper formally defines the problem of probabilistic representative frequent pattern (P-RFP) mining, which aims to find the minimal set of patterns with sufficiently high probability to represent all other patterns. The problem's bottleneck turns out to be checking whether a pattern can probabilistically represent another, which involves the computation of a joint probability of supports of two patterns. To address the problem, we propose a novel and efficient dynamic programming-based approach. Moreover, we have devised a set of effective optimization strategies to further improve the computation efficiency. Our experimental results demonstrate that the proposed P-RFP mining effectively reduces the size of probabilistic frequent patterns. Our proposed approach not only discovers the set of P-RFPs efficiently, but also restores the frequency probability information of patterns with an error guarantee.
Hasan, M., Xu, M., He, S. & Chen, L. 2012, 'Shot Classification Using Domain Specific Features for Movie Management', Lecture Notes in Computer Science, Springer, Busan, South Korea, pp. 314-318.
View/Download from: UTS OPUS or Publisher's site
Among many video types, movie content indexing and retrieval is a significantly challenging task because of the wide variety of shooting techniques and the broad range of genres. A movie consists of a series of video shots. Managing a movie at shot level provides a feasible way for movie understanding and summarization. Consequently, an effective shot classification is greatly desired for advanced movie management. In this demo, we explore novel domain specific features for effective shot classification. Experimental results show that the proposed method classifies movie shots from wide range of movie genres with improved accuracy compared to existing work
Long, G., Chen, L., Zhu, X. & Zhang, C. 2012, 'TCSST: transfer classification of short & sparse text using external data', Proc. Of The 21st ACM Conference on Information and Knowledge Management (CIKM-12), ACM Conference on Information and Knowledge Management, ACM, Hawaii, USA, pp. 764-772.
View/Download from: UTS OPUS or Publisher's site
Short & sparse text is becoming more prevalent on the web, such as search snippets, micro-blogs and product reviews. Accurately classifying short & sparse text has emerged as an important while challenging task. Existing work has considered utilizing external data (e.g. Wikipedia) to alleviate data sparseness, by appending topics detected from external data as new features. However, training a classifier on features concatenated from different spaces is not easy considering the features have different physical meanings and different significance to the classification task. Moreover, it exacerbates the "curse of dimensionality" problem. In this study, we propose a transfer classification method, TCSST, to exploit the external data to tackle the data sparsity issue. The transfer classifier will be learned in the original feature space. Considering that the labels of the external data may not be readily available or sufficiently enough, TCSST further exploits the unlabeled external data to aid the transfer classification. We develop novel strategies to allow TCSST to iteratively select high quality unlabeled external data to help with the classification. We evaluate the performance of TCSST on both benchmark as well as real-world data sets. Our experimental results demonstrate that the proposed method is effective in classifying very short & sparse text, consistently outperforming existing and baseline methods
Chen, L. & Zhang, C. 2011, 'Semi-supervised Variable Weighting for Clustering', Proceedings of the Eleventh SIAM International Conference on Data Mining, SDM, SIAM / Omnipress, Mesa, Arizona, USA, pp. 863-871.
View/Download from: UTS OPUS
Semi-supervised learning, which uses a small amount of labeled data in conjunction with a large amount of unlabeled data for training, has recently attracted huge research attention due to the considerable improvement in learning accuracy. In this work, we focus on semi- supervised variable weighting for clustering, which is a critical step in clustering as it is known that interesting clustering structure usually occurs in a subspace defined by a subset of variables. Besides exploiting both labeled and unlabeled data to effectively identify the real importance of variables, our method embeds variable weighting in the process of semi-supervised clustering, rather than calculating variable weights separately, to ensure the computation efficiency. Our experiments carried out on both synthetic and real data demonstrate that semi-supervised variable weighting signicantly improves the clustering accuracy of existing semi-supervised k-means without variable weighting, or with unsupervised variable weighting.
Papapetrou, O. & Chen, L. 2011, 'XStreamCluster: an Efficient Algorithm for Streaming XML Data Clustering', Lecture Notes in Computer Science: Database Systems for Advanced Applications 16th International Conference, DASFAA 2011, International Conference on Database Systems for Advanced Applications, Springer, Hongkong, China, pp. 496-510.
View/Download from: UTS OPUS or Publisher's site
XML clustering finds many applications, ranging from storage to query processing. However, existing clustering algorithms focus on static XML collections, whereas modern information systems frequently deal with streaming XML data that needs to be processed online. Streaming XML clustering is a challenging task because of the high computational and space efficiency requirements implicated for online approaches. In this paper we propose XStreamCluster, which addresses the two challenges using a two-layered optimization. The bottom layer employs Bloom filters to encode the XML documents, providing a space-efficient solution to memory usage. The top layer is based on Locality Sensitive Hashing and contributes to the computational efficiency. The theoretical analysis shows that the approximate solution of XStreamCluster generates similarly good clusters as the exact solution, with high probability. The experimental results demonstrate that XStreamCluster improves both memory efficiency and computational time by at least an order of magnitude without affecting clustering quality, compared to its variants and a baseline approach.
Yu, P.S., Fan, W., Nejdl, W., Chen, L., Sun, A., Simovici, D., Baralis, E., Nguifo, E.M., Xu, G., Yin, J., Ceci, M., Cortez, P., Christen, P., Berka, P., Alves, R., Xu, S., Elomaa, T., Kosters, W., Graco, W., Wang, W., Balke, W.T. & Zhao, Y. 2011, 'Preface', Proceedings - IEEE International Conference on Data Mining, ICDM.
View/Download from: Publisher's site
Su, H., Chen, L., Ye, Y., Sun, Z. & Wu, Q. 2010, 'A refinement approach to handling model misfit in semi-supervised learning', Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), Springer, Chongqing, China, pp. 75-86.
View/Download from: UTS OPUS or Publisher's site
Semi-supervised learning has been the focus of machine learning and data mining research in the past few years. Various algorithms and techniques have been proposed, from generative models to graph-based algorithms. In this work, we focus on the Cluster-
Xu, M., Chen, L., He, S., Xu, C. & Jin, J. 2010, 'Adaptive Local Hyperplanes for MTV affective analysis', Proceedings of the 2nd International Conference on Internet Multimedia Computing and Service, ICIMCS'10, International Conference on Internet Multimedia Computing and Service, ACM, Harbin, China, pp. 167-170.
View/Download from: UTS OPUS or Publisher's site
Affective analysis attracts increasing attention in multimedia domain since affective factors directly reflect audiences' attention, evaluation and memory. Existing study focuses on mapping low-level affective features to high-level emotions by applying

Journal articles

Wang, H., Wu, J., Pan, S., Zhang, P. & Chen, L. 2017, 'Towards large-scale social networks with online diffusion provenance detection', Computer Networks, vol. 114, pp. 154-166.
View/Download from: Publisher's site
© 2016 Elsevier B.V.In this paper we study a new problem of online discovering diffusion provenances in large networks. Existing work on network diffusion provenance identification focuses on offline learning where data collected from network detectors are static and a snapshot of the network is available before learning. However, an offline learning model does not meet the need for early warning, real-time awareness, or a real-time response to malicious information spreading in networks. To this end, we propose an online regression model for real-time diffusion provenance identification. Specifically, we first use offline collected network cascades to infer the edge transmission weights, and then use an online l 1 non-convex regression model as the identification model. The proposed methods are empirically evaluated on both synthetic and real-world networks. Experimental results demonstrate the effectiveness of the proposed model.
Wang, H., Zhang, P., Zhu, X., Tsang, I.W.H., Chen, L., Zhang, C. & Wu, X. 2017, 'Incremental Subgraph Feature Selection for Graph Classification', IEEE Transactions on Knowledge and Data Engineering, vol. 29, no. 1, pp. 128-142.
View/Download from: Publisher's site
Wang, W., Yin, H., Chen, L., Sun, Y., Sadiq, S. & Zhou, X. 2017, 'ST-SAGE: A spatial-Temporal sparse additive generative model for spatial item recommendation', ACM Transactions on Intelligent Systems and Technology, vol. 8, no. 3.
View/Download from: Publisher's site
©c 2017 ACM 2157-6904/2017/04-ART48 15.00. With the rapid development of location-based social networks (LBSNs), spatial item recommendation has become an important mobile application, especially when users travel away from home. However, this type of recommendation is very challenging compared to traditional recommender systems. A user may visit only a limited number of spatial items, leading to a very sparse user-item matrix. This matrix becomes even sparser when the user travels to a distant place, as most of the items visited by a user are usually located within a short distance from the user's home. Moreover, user interests and behavior patterns may vary dramatically across different time and geographical regions. In light of this, we propose ST-SAGE, a spatial-Temporal sparse additive generative model for spatial item recommendation in this article. ST-SAGE considers both personal interests of the users and the preferences of the crowd in the target region at the given time by exploiting both the co-occurrence patterns and content of spatial items. To further alleviate the data-sparsity issue, ST-SAGE exploits the geographical correlation by smoothing the crowd's preferences over a well-designed spatial index structure called the spatial pyramid. To speed up the training process of ST-SAGE, we implement a parallel version of themodel inference algorithm on the GraphLab framework.We conduct extensive experiments; the experimental results clearly demonstrate that ST-SAGE outperforms the state-of-The-Art recommender systems in terms of recommendation effectiveness, model training efficiency, and online recommendation efficiency.
Hou, S., Chen, L., Tao, D., Zhou, S., Liu, W. & Zheng, Y. 2017, 'Multi-layer multi-view topic model for classifying advertising video', Pattern Recognition, vol. 68, pp. 66-81.
View/Download from: Publisher's site
© 2017 Elsevier Ltd The recent proliferation of advertising (ad) videos has driven the research in multiple applications, ranging from video analysis to video indexing and retrieval. Among them, classifying ad video is a key task because it allows automatic organization of videos according to categories or genres, and this further enables ad video indexing and retrieval. However, classifying ad video is challenging compared to other types of video classification because of its unconstrained content. While many studies focus on embedding ads relevant to videos, to our knowledge, few focus on ad video classification. In order to classify ad video, this paper proposes a novel ad video representation that aims to sufficiently capture the latent semantics of video content from multiple views in an unsupervised manner. In particular, we represent ad videos from four views, including bag-of-feature (BOF), vector of locally aggregated descriptors (VLAD), fisher vector (FV) and object bank (OB). We then devise a multi-layer multi-view topic model, mlmv_LDA, which models the topics of videos from different views. A topical representation for video, supporting category-related task, is finally achieved by the proposed method. Our empirical classification results on 10,111 real-world ad videos demonstrate that the proposed approach effectively differentiate ad videos.
Hou, S., Zhou, S., Chen, L., Feng, Y. & Awudu, K. 2016, 'Multi-label learning with label relevance in advertising video', Neurocomputing, vol. 171, pp. 932-948.
View/Download from: Publisher's site
The recent proliferation of videos has brought out the need for applications such as automatic annotation and organization. These applications could greatly benefit from the respective thematic content depending on the type of video. Unlike the other kinds of video, an advertising video usually conveys a specific theme in a certain time period (e.g. drawing the audiences attention to a product or emphasizing the brand). Traditional multi-label algorithms may not work effectively with advertising videos due mainly to their heterogeneous nature. In this paper, we propose a new learning paradigm to resolve the problems arising out of traditional multi-label learning in advertising videos through label relevance. Aiming to address the issue of label relevance, we firstly assign each label with label degree (LD) to classify all the labels into three groups such as first label (FL), important label (IL) and common label (CL), and then propose a Directed Probability Label Graph (DPLG) model to mine the most related labels from the multi-label data with label relevance, in which the interdependency between labels is considered. In the implementation of DPLG, the labels that appear occasionally and possess inconspicuous co-occurrences are consequently eliminated effectively, employing -filtering and -pruning processes, respectively. And then the graph theory is utilized in DPLG to acquire Correlative Label-Sets (CLSs). Lastly, the searched Correlative Label-Sets (CLSs) are utilized to enhance multi-label annotation. Experimental results on advertising videos and several publicly available datasets demonstrate the effectiveness of the proposed method for multi-label annotation with label relevance
Zhang, Y., Wu, J., Cai, Z., Zhang, P. & Chen, L. 2016, 'Memetic Extreme Learning Machine', Pattern Recognition, vol. 58, pp. 135-148.
View/Download from: Publisher's site
© 2016. Extreme Learning Machine (ELM) is a promising model for training single-hidden layer feedforward networks (SLFNs) and has been widely used for classification. However, ELM faces the challenge of arbitrarily selected parameters, e.g., the network weights and hidden biases. Therefore, many efforts have been made to enhance the performance of ELM, such as using evolutionary algorithms to explore promising areas of the solution space. Although evolutionary algorithms can explore promising areas of the solution space, they are not able to locate global optimum efficiently. In this paper, we present a new Memetic Algorithm (MA)-based Extreme Learning Machine (M-ELM for short). M-ELM embeds the local search strategy into the global optimization framework to obtain optimal network parameters. Experiments and comparisons on 46 UCI data sets validate the performance of M-ELM. The corresponding results demonstrate that M-ELM significantly outperforms state-of-the-art ELM algorithms.
Yin, H., Cui, B., Chen, L., Hu, Z. & Zhang, C. 2015, 'Modeling Location-Based User Rating Profiles for Personalized Recommendation', ACM Transactions on Knowledge Discovery from Data, vol. 9, no. 3, pp. 1-41.
View/Download from: Publisher's site
Yin, H., Cui, B., Chen, L., Hu, Z. & Zhou, X. 2015, 'Dynamic User Modeling in Social Media Systems', ACM Transactions on Information Systems, vol. 33, no. 3, pp. 1-44.
View/Download from: Publisher's site
Yin, H., Cui, B., Sun, Y., Hu, Z. & Chen, L. 2014, 'LCARS: A Spatial Item Recommender System', ACM Trans. Inf. Syst., vol. 32, no. 3, pp. 1-37.
View/Download from: Publisher's site
Newly emerging location-based and event-based social network services provide us with a new platform to understand users' preferences based on their activity history. A user can only visit a limited number of venues/events and most of them are within a limited distance range, so the user-item matrix is very sparse, which creates a big challenge to the traditional collaborative filtering-based recommender systems. The problem becomes even more challenging when people travel to a new city where they have no activity information. In this article, we propose LCARS, a location-content-aware recommender system that offers a particular user a set of venues (e.g., restaurants and shopping malls) or events (e.g., concerts and exhibitions) by giving consideration to both personal interest and local preference. This recommender system can facilitate people's travel not only near the area in which they live, but also in a city that is new to them. Specifically, LCARS consists of two components: offline modeling and online recommendation. The offline modeling part, called LCA-LDA, is designed to learn the interest of each individual user and the local preference of each individual city by capturing item cooccurrence patterns and exploiting item contents. The online recommendation part takes a querying user along with a querying city as input, and automatically combines the learned interest of the querying user and the local preference of the querying city to produce the top-k recommendations. To speed up the online process, a scalable query processing technique is developed by extending both the Threshold Algorithm (TA) and TA-approximation algorithm. We evaluate the performance of our recommender system on two real datasets, that is, DoubanEvent and Foursquare, and one large-scale synthetic dataset. The results show the superiority of LCARS in recommending spatial items for users, especially when traveling to new cities, in terms of both effectiveness and efficiency. Beside...
Li, B., Chen, L., Zhu, X. & Zhang, C. 2013, 'Noisy but Non-malicious User Detection in Social Recommender Systems', World Wide Web, vol. 16, no. 5-6, pp. 677-699.
View/Download from: UTS OPUS or Publisher's site
Social recommender systems largely rely on user-contributed data to infer users' preference. While this feature has enabled many interesting applications in social networking services, it also introduces unreliability to recommenders as users are allowed to insert data freely. Although detecting malicious attacks from social spammers has been studied for years, little work was done for detecting Noisy but Non-Malicious Users (NNMUs), which refers to those genuine users who may provide some untruthful data due to their imperfect behaviors. Unlike colluded malicious attacks that can be detected by finding similarly-behaved user profiles, NNMUs are more difficult to identify since their profiles are neither similar nor correlated from one another. In this article, we study how to detect NNMUs in social recommender systems. Based on the assumption that the ratings provided by a same user on closely correlated items should have similar scores, we propose an effective method for NNMU detection by capturing and accumulating userâs âself-contradictionsâ, i.e., the cases that a user provides very different rating scores on closely correlated items. We show that self-contradiction capturing can be formulated as a constrained quadratic optimization problem w.r.t. a set of slack variables, which can be further used to quantify the underlying noise in each test user profile. We adopt three real-world data sets to empirically test the proposed method. The experimental results show that our method (i) is effective in real-world NNMU detection scenarios, (ii) can significantly outperform other noisy-user detection methods, and (iii) can improve recommendation performance for other users after removing detected NNMUs from the recommender system.