Whereas a lot of DC homeworks are group homeworks, some (such as this one) are individual homeworks. This means, each individual submits the homework and no consultation with others is permitted. One survey-based individual homework was already putup.
Recall the subject of latent topic-mining we'd covered in DC Session 4. The machine "discovers" topics latent in text corpora and assigns probabilities for topic membership of word-tokens and for topic-presence in documents. This survey aims to assess machine classification of phrase-tokens and 'documents' when pitted against a superior human standard of classification.
There are no right or wrong answers in what follows, so just diligently solve the questions posed to you. This homework is graded on timeliness and completeness.
We've test-run the surveys. I think most people should be able to complete it comfortably within 15-20 minutes. Hence, a relatively deadline is given.
This is the survey link for Section A people.
This one is for Section B people.
Deadline for filling up: By midnight Sunday 04-October. Any queries, clarifications etc, contact me.
Sudhir
Hi Prof,
ReplyDeleteThere is no back button on the survey pages. I skipped one to make sure my understanding was correct, and now I can't go back. :)
I can try clearing cookies and trying again, but just wanted to check if this is by design and If i already lost some points.
Thanks
Ankur - Batch B
We didn't build a back button, true. Pls clear the cookies cache and retry. Thanks.
DeleteSudhir