I wish to automate the process of users moderation on a social network. My interest is to automatically prevent the use of abusive and erotic words on a discussion forum. Can someone help me on this? Thanks in advance.
The question has great importance in social network as we see that most of the people post comment which contains vulgar and abusive contents. Also, it is very important to find and remove such contents accordingly. For that, i would suggest that those terms should be identified which are most commonly used in abusive language. Naive Bayes classifier can categorize such comments either abusive or non abusive. There can be other solutions such as those people who are using abusive language upon is warned with a message and a threshold should be fixed for blocking/ warning such a person.
Hope it help..
Just like spam filter, you can do it by using proper supervised classification algorithm that can filter such comments and declare it as abusive. The later action depend upon you whether you are going to block or warn that particular person. Some effort would be required to find features and terms that can be used for the identification of such contents.
I would also add just be careful to not filter out legitimate terminology along with crude or offensive terms. For example in the UK we have town names - Clitheroe and Scunthorpe- which are often the victims of filters.
Well I suppose a blacklist-whitelist approach may work when setting up your initial algorithm. But really as Zafar Ali pointed out designing a good supervised classification algorithm will work too. I guess it depends more on how you intend to implement the automation. Good prior design is what I am trying to stress here.
- Badges
- Science topic
- Similar topics
- Thesis Research