Murata gave a talk (title: "Deep Learning Approaches for Eruption Prediction of Sakurajima") as an invited speaker of 115th SIG-KBS meeting of JSAI. (2018.11.23)
Mr. Hiep Le's work of the prediction of volcanic eruption of Sakurajima appeared in an article of monthly magazine Fole (in Japanese) issued by Mizuho Research Institute in October 2018. (2018.9.27)
A paper (Title:"Learning Community Structure with Variational Autoencoder") written by Mr. Choong Jun Jin (second grade of doctoral course student) and Dr. Liu Xin (AIRC, AIST) is accepted as a full paper of IEEE ICDM 2018 (acceptance rate:8.86%). (2018.8.20)
Murata taught an intensive lecture on Complex Networks at Shanghai University of Finance and Economics. (2018.7.12-17)
Mr. Hiep Le's paper (Title: "Deep Modular Multimodal Fusion on Multiple Sensors for Volcano Activity Recognition") has been accepted by ECML-PKDD 2018 (acceptance rate:27%). (2018.6.19)
Mr. Hiep Le's research on predicting volcanic eruptions of Sakurajima (presented at JSAI 2018 on June 6, 2018) appeared in a local TV news of NHK Kagoshima and in an atricle of South Japan Newspaper (2018.6.7)
"Kagoshima NEWS WEB -- AI predicts eruptions of Sakurajima" (NHK Kagoshima, in Japanese)
"AI predicts eruptions of Sakurajima --- 51.9% accuracy achieved with four-stage warning system by research team of Tokyo Tech & Kyoto Univ." (South Japan Newspaper Company, in Japanese)
An article about Mr. Hiep Le's work on Sakurajima volcanic activity recognition appeared in Nikkei Newspaper (page 9) on Jan. 15, 2018. (2018.1.15)
"Tokyo Tech & Kyoto Univ. use AI for volcano activity recognition (Nikkei Newspaper (written in Japanese))" "Deep Learning Approach for Volcano Activity Recognition (News of the Dept. of CS (written in Japanese))" "Deep Learning Approach for Volcano Activity Recognition (link to an article of KBS meeting in Nov. 2017(written in Japanese))"
call for postdocs: Our laboratory takes part in RWBC-OIL(AIST-Tokyo Tech Real World Big-Data Computation Open Innovation Laboratory).
RWBC-OIL is a joint laboratory of Tokyo Tech and AIST for the analysis of real-world big data started from Feb. 2017 (news from Tokyo Tech (in Japanese), news from AIST (in Japanese)).
RWBC-OIL hires researchers of machine learning / deep learning / graph mining / high performance computing. Please read call for postdocs (in Japanese) for more information. If you are interested in, please contact Murata. (2017.4.6)
Mr. Yohei Oonuki receives JSAI Incentive Award for his presentation (title: "Prediction of relations among RDF entities by DNN" (in Japanese)) at the 41st SIG meeting on Semantic Web and Ontology. (2017.5.8)
Mr. Shun Nukui received JSAI Annual Conference Student Incentive Award for his presentation (title: "Constrained community detection based on semi-supervised non-negative matrix factorization on TensorFlow" (in Japanese)) at 2016 JSAI Annual Conference. (2017.3.1)
Murata was invited to MARAMI 2016 as an invited speaker, but the talk was canceled. (2016.10.14)
Murata gave a talk ("Mining and learning on heterogeneous networks") as an invited speaker at WIMS 2016 (slides). (2016.6.15)
Dr. Xin Liu and Mr. Weichu Liu's paper (title: "Community Detection in Multi-Partite Multi-Relational Networks Based on Information Compression") was accepted by New Generation Computing. (2016.3.7)
Mr. Phiradet Bangcharoensap's paper (title: "Transductive Classification on Heterogeneous Information Networks with Edge Betweenness-based Normalization") was accepted by WSDM 2016 conference (acceptance rate: 18.2%). (2015.10.14)
Mr. Phiradet Bangcharoensap's paper (title: "Two step graph-based semi-supervised learning for online auction fraud detection") was accepted by ECML-PKDD 2015 conference (acceptance rate: 28.6%). (2015.6.3)
Mr. Keisuke Nakata presented his work at CompleNet 2015(the 6th Workshop on Complex Networks). His photo is here. (2015.3.25)
Mr. Keisuke Nakata (title: "Speeding-up of Constrained Community Detection and Development of its Interactive Environment") received JSAI Annual Conference Award. 19 papers were selected out of 584 papers. (2014.7.22)
Dr. Xin Liu and Mr. Weichu Liu's paper (A Framework for Community Detection in Heterogeneous Multi-Relational Networks) was accepted by Advances in Complex Systems.(2014.7.9)
Dr. Xin Liu's paper (Detecting network communities beyond assortativity-related attributes) was accepted by Physical Review E.(2014.6.19)
Mr. Ikematsu's programs for detecting communities from n-partite networks are available online.
Reference:"Improvement of a Tripartite Modularity and Its Optimization Method"(written in Japanese)
(Transactions of the Japanese Society for Artificial Intelligence, Vol.29, No.2, pp.245-258, 2014)(2014.2.20)