Abstract: Network motif discovery, community detection, and frequent subgraph mining require subgraph enumeration. Recent works parallelize subgraph enumeration using GPUs to accelerate execution. Set intersection operations take up to 95% of the processing time in…
Abstract: Big data applications generate massive amounts of high-dimensional, real-time, streaming data. These applications require effective and efficient data stream clustering. Well-known data stream clustering algorithms based on the popular online-offline framework still face major…
Abstract: Time-series shapelets, discriminative subsequences, are effective for time series classification (tsc). Tsc accuracy depends on shapelet quality. However, major research has focused on accurate models from some shapelet candidates. Existing studies use simple methods…
Abstract: Social media and bibliographic datasets can be graphed. Clustering graphs reveals data structure. Node attributes in attributed graphs can improve clustering. Attributed graph clustering methods typically separate structural and attribute similarity. This paper models…
Abstract: HINs are used to model information systems with multi-type objects and relations. Homogeneous graphs have one type of nodes and edges. Web search, link prediction, and clustering require measuring object similarities. HIN similarity measures…
Abstract: Heterogeneous information networks (HINs)—typed graphs with labeled nodes and edges—are popular in academia and industry. We discover the k most important meta paths in real time for friend search, product recommendation, anomaly detection, and…
Abstract: Matrix decomposition is essential for mining big data from modern applications. However, processing very large data on a single machine is inefficient or impossible. Big data are also distributed across multiple machines. Thus, heterogeneous…
Abstract: Several discretization algorithms have been developed, but efficient, accurate discretization remains a challenge. This paper proposes SPID5, a new discretization algorithm based on two supervised heuristics, one local and one global, that synergize. The…
Abstract: Temporal point process (TPP) modeling is expressive. TPP modeling rarely investigates event sequence clustering temporal patterns. Reinforcement learning assumes the observed sequences are generated by a mixture of latent policies to solve this problem….
Abstract: Data analysis involves community detection. Algorithms to solve this problem have been proposed for decades. Most community detection research ignores statistical significance. Although statistically significant communities have been mined, deriving an analytical solution of…
Abstract: Spatio-temporal data is increasingly available due to the rapid development of GPS, mobile devices, and remote sensing. Human mobility understanding, smart transportation, urban planning, public safety, health care, and environmental management depend on spatio-temporal…
Abstract: Adverse event detection is essential for identifying product defects, disasters, and major socio-political events. Adverse drug events cause many hospitalizations and deaths annually. Search query logs are a key detection channel because users start…
WhatsApp us