Research » Web Spam Detection » Datasets » WEBSPAM-UK2007 » Assessment comments


During the assessment of the dataset, the assesors were allowed to leave comments on the hosts they were evaluating, explaining why they were giving a borderline, spam, or nonspam vote; or commenting on other user's comments if they were relabelling. Comments were left only for a small portion of the data (980 hosts in total), including hosts in all the classes (borderline, spam, and nonspam).

To download this file, if you have not done it yet, please sign and fax the agreement for the contents of the collection, and then use your password to access the link below:

This file has been anonymized according to the privacy policy.

For inquiries contact Carlos Castillo