October 19, 2016
2:00 pm - 5:30 pm
Hall C

Track: General
Type: Posters
Level: All

Building a classifier which gives high accuracy does not always suffice. We dig deep into the training URL dataset to better understand how the classifier detects or learns from the labeled data provided. We analyze the nature of the URLs used based on different aspects like the presence of common country codes, server side language extension markers, special strings, etc for URL classification.

Speaker(s)