Although our main focus is on academic documents in AI, data mining and machine learning fields, academic documents from different fields can also be found in the system based on their citation relationship with one of the papers in the conferences or journals we are working on.
We focus on academic documents published in 17 different conferences and journals between 2001 to 2015, including ICDE, KDD, TKDE, VLDB, CIKM, NIPS, ICML, ICDM, PKDD, SDM, WSDM, AAAI, IJCAI, DMKD, WWW, KAIS and TKDD. Using the Microsoft graph dataset, we then extended these documents, adding their references and the references of their references (up to 2 hops away).
Search for the paper, and select it. Go down to the citation graph and check for the darker nodes. Clicking on the darker nodes will display the information about the clicked item. If dataset weblinks was found in the paper, clicking on the "link" tab would display the weblinks and clicking on the weblinks would take the user you the dataset webpage.
This can be due to several reasons: - The paper did not provide clear information on the datasets used in their evaluation. - The paper used closed source datasets which are not available to the public. - We were unable to detect the dataset information.