A group of researchers are building a natural language processing to detect hate speech online
Called ‘Project Angel’, the project began after the group, found incidences of depression among 25-30% of girls in that age group
Hyderabad: In a bid to help adolescent girls combat online social biases or toxicity in the form of body-shaming, sexual harassment, etc a group of researchers from the International Institute of Technology-Hyderabad (IIIT-H) are building a natural language processing (NLP) tool which can detect such hate speech online. Called ‘Project Angel’, the project began after the group, through interviews and surveys, found incidences of depression among 25-30% of girls in that age group.
Resolving to help the troubled teenagers, researchers at IIIT-H’s Information Retrieval and Extraction Lab led by Prof. Vasudeva Varma, developed deep neural networks that not only detected online sexist comments but also automatically labelled and categorized them. The team was able to come up with new insights not just from a technological point of view, but also from a social sciences perspective, that will assist social scientists and policy makers to study and counter sexism better.
“The NLP was being built to detect hate-speech was ongoing from three to four years ago, but we started working on Project Angel about a year ago after interviewing hundreds of adolescent girls. The technology is specifically being tailored for teenage girls as that is a more vulnerable age," Prof. Varma told Mint.
He added that the NLP being built to help adolescent girls is not intrusive, and that they are aware that teenagers will not want to compromise on their privacy. NLP entails applying algorithms to identify and extract the natural language rules, such that unstructured language data is converted into a form that computers can understand. This particular project involves semantics, which refers to the meaning that is conveyed by a text.
Prof. Varma said that all one has to do, for Project Angel to be of help, is to be on its network. “For example, if you are on Twitter, all of you have to do is follow our NLP. It is an intelligent listener with which you can interact. Project Angel is like another person, who can read what is on your wall, and if it detects some hate-speech, it will alert you about it," he added.
As part of the project, the IIT-H team is also building several building programming the NLP to understand body-shaming specifically. “It is one of the main aspects through which their emotional state (of a girl) gets disturbed. It leads to formation of relationships they don’t want to stay in. We are in stage 2 currently, and in stage 3 of Project Angel will connect you with other people who can potentially help you," informed Prof. Varma.
The team is also working on finding the right point of interventions necessary, to bring in the required positivity into any of the platforms used by teens through the project.