Rose(Qi) Yu

Processed ACM dataset for group anomaly detecton


This dataset contains the ACM dataset we used for group anomaly detection in the dynamic setting. The raw data we get is from ArnetMiner , an academic scholar social network search engine. The dataset contains a ten year coo-authorship network of 4474 scholars and their authored paper abstracts ( in Bag of Words form) from year 2000 to year 2009. The number of key words we used for experiments is 8025.


Processed ACM dataset(Zip file 799 KB),