Report

Multi-label Relational Neighbor Classification using Social Context Features Xi Wang and Gita Sukthankar Department of EECS University of Central Florida Motivation The conventional relational classification model focuses on the single-label classification problem. Real-world relational datasets contain instances associated with multiple labels. Connections between instances in multi-label networks are driven by various casual reasons. Artificial Intelligence Data Mining Machine Learning Example: Scientific collaboration network 1 Problem Formulation Node classification in multi-relational networks Input: Network structure (i.e., connectivity information) Labels of some actors in the network Output: Labels of the other actors 2 Classification in Networked Data Homophily: nodes with similar labels are more likely to be connected Markov assumption: The label of one node depends on that of its immediate neighbors in the graph Relational models are built based on the labels of neighbors. Predictions are made using collective inference. 3 Contribution A new multi-label iterative relational neighbor classifier (SCRN) Extract social context features using edge clustering to represent a node’s potential group membership Use of social features boosts classification performance over benchmarks on several real-world collaborative networked datasets 4 Relational Neighbor Classifier The Relational Neighbor (RN) classifier proposed by Macskassy et al. (MRDM’03), is a simple relational probabilistic model that makes predictions for a given node based solely on the class labels of its neighbors. Training Graph Iteration 1 Iteration 2 5 Relational Neighbor Classifier Weighted-vote relational neighbor classifier (wvRN) estimates prediction probability as: P ( Li c | v i ) 1 z w (v , v i j ) P(L j c | N j ) v jN i Here z is the usual normalization factor, and w(vi , v j ) is the weight of the link between node vi and v j 6 Apply RN in Multi-relational Network : nodes with both labels (red, green) : nodes with green label only : nodes with red label only Ground truth 7 Edge-Based Social Feature Extraction Connections in human networks are mainly affiliationdriven. Since each connection can often be regarded as principally resulting from one affiliation, links possess a strong correlation with a single affiliation class. The edge class information is not readily available in most social media datasets, but an unsupervised clustering algorithm can be applied to partition the edges into disjoint sets (KDD’09,CIKM’09). 8 Cluster edges using K-Means Scalable edge clustering method proposed by Tang et al. (CIKM’09). Each edge is represented in a feature-based format, where each edge is characterized by its adjacent nodes. K-means clustering is used to separate the edges into groups, and the social feature (SF) vector is constructed based on edge cluster IDs. Original network Step1 : Edge representations Step2: Construct social features 9 Edge-Clustering Visualization Figure: A subset of DBLP with 95 instances. Edges are clustered into 10 groups, with each shown in a different color. 10 Proposed Method: SCRN The initial set of reference features for class c can be defined as the weighted sum of social feature vectors for nodes known to be in class c: RV (c) = 1 c P(l å i =1)´ SF(vi ) K | Vc | vi ÎVcK Then node vi’s class propagation probability for class c conditioned on its social features: PCP (lic | SF(vi )) = sim(SF(vi ), RV(c)) 11 SCRN SCRN estimates the class-membership probability of node vi belonging to class c using the following equation: 1 P(l | Ni , SF(vi )) = å PCP (lic | SF(vi ))´ w(vi , v j ) ´ P(l cj | N j ) z v j ÎNi c i class propagation probability similarity between connected nodes (link weight) class probability of its neighbors 12 SCRN Overview Input: {G,V, E,C, LK } , Max_Iter Output: LU for nodes in V U 1. Construct nodes’ social feature space 2. Initialize the class reference vectors for each class 3. Calculate the class-propagation probability for each test node 4. Repeat until # of iterations > Max_Iter or predictions converge Estimate test node’s class probability Update the test node’s class probability in collective inference Update the class reference vectors Re-calculate each node’s class-propagation probability 13 SCRN Visualization Figure: SCRN on synthetic multi-label network with 1000 nodes and 32 classes (15 iterations). 14 Datasets DBLP We construct a weighted collaboration network for authors who have published at least 2 papers during the 2000 to 2010 time- frame. We selected 15 representative conferences in 6 research areas: DataBase: ICDE,VLDB, PODS, EDBT Data Mining: KDD, ICDM, SDM, PAKDD Artificial Intelligence: IJCAI, AAAI Information Retrieval: SIGIR, ECIR Computer Vision: CVPR Machine Learning: ICML, ECML 15 Datasets IMDb We extract movies and TV shows released between 2000 and 2010, and those directed by the same director are linked together. We only retain movies and TV programs with greater than 5 links. Each movie can be assigned to a subset of 27 different candidate movie genres in the database such as “Drama", “Comedy", “Documentary" and “Action”. 16 Datasets YouTube A subset of data (15000 nodes) from the original YouTube dataset[1] using snowball sampling. Each user in YouTube can subscribe to different interest groups and add other users as his/her contacts. Class labels are 47 interest groups. [1] http://www.public.asu.edu/~ltang9/social_ dimension.html 17 Comparative Methods Edge (EdgeCluster) wvRN Prior Random 18 Experiment Setting Size of social feature space : 1000 for DBLP and YouTube; 10000 for IMDb Class propagation probability is calculated with the Generalized Histogram Intersection Kernel. Relaxation Labeling is used in the collective inference framework for SCRN and wvRN. We assume the number of labels for testing nodes is known. 19 Experiment Setting We employ the network cross-validation (NCV) method (KAIS’11) to reduce the overlap between test samples. Classification performance is evaluated based on Micro-F1, Macro-F1 and Hamming Loss. 20 Results (Micro-F1) DBLP Micro-F1 accuracy (%) 70 60 SCRN 50 Edge 40 wvRN 30 Prior 20 Random 10 5 10 15 20 25 Training data percentage(%) 30 21 Results (Macro-F1) DBLP Macro-F1 accuracy (%) 70 60 SCRN 50 Edge 40 wvRN 30 Prior 20 Random 10 5 10 15 20 25 Training data percentage (%) 30 22 Results (Hamming Loss) DBLP 23 Results (Hamming Loss) IMDb 24 Results (Hamming Loss) YouTube 25 Conclusion Links in multi-relational networks are heterogeneous. SCRN exploits label homophily while simultaneously leveraging social feature similarity through the introduction of class propagation probabilities. Significantly boosts classification performance on multilabel collaboration networks. Our open-source implementation of SCRN is available at: http://code.google.com/p/multilabel-classification-on-social-network/ 26 Reference MACSKASSY, S. A., AND PROVOST, F. A simple relational classifier. In Proceedings of the Second Workshop on Multi-Relational Data Mining (MRDM) at KDD, 2003, pp. 64–76. TANG, L., AND LIU, H. Relational learning via latent social dimensions. In Proceedings of the ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (KDD), 2009, pp. 817–826. TANG, L., AND LIU, H. Scalable learning of collective behavior based on sparse social dimensions. In Proceedings of International Conference on Information and Knowledge Management (CIKM), 2009, pp. 1107-1116. NEVILLE, J., GALLAGHER, B., ELIASSI-RAD, T., AND WANG, T. Correcting evaluation bias of relational classifiers with network cross validation. Knowledge and Information Systems (KAIS), 2011, pp. 1–25. 27 Thank you! 28