Data Analytics for Structured Data

Since the early 2000s, we have been recognized for research accomplishments in data analytics for structured data. Of particular importance is the following work.

• We proposed the celebrated first algorithm called AGM (a-priori-based graph mining), which is the one that opened a new door to the fertile field of graph mining.

  Akihiro Inokuchi, Takashi Washio, Hiroshi Motoda, "An Apriori-Based Algorithm for Mining Frequent Substructures from Graph Data," PKDD 2000: 13-23.

• We first proposed the notion of graph kernels, which enables us to computed a kernel function for structured data.

  Hisashi Kashima, Koji Tsuda, Akihiro Inokuchi, "Marginalized Kernels Between Labeled Graphs," ICML 2003: 321-328

In addition to these, we have been actively studied data analytics for structured data, with particular emphasis on the application to the real world. Examples include:

• Online anomaly detection from time-evolving graphs (Ide and Kashima, KDD 2004)
  
  
• Induced ordered tree mining in tree-structured databases (Hido and Kawano, ICDM 2005)
  
  
• Supervised link prediction for network data (Kashima and Abe, ICDM 2006).

Japanese page includes more detailed information.

Members

• Akira Tajima (manager)
  
  
• Hisashi Kashima
  
  
• Tsuyoshi Ide
  
  
• Rikiya Takahashi
  
  
• Toshihiro Takahashi
  
  
• Shoko Suzuki
  
  
• Shohei Hido
  
  

Links

• Publication list
  
  
• Knowledge discovery & data mining research in IBM
  
  
• Services Sciences, Management and Engineering in IBM
  

Past projects

• Data Mining
  
  
• Spatial Data Mining