Transphobic social detector

Context

Part of Maavarim activities are to be the trans community representatives in the Knesset. They are forced to back their opinions with data.

For example: we have a strong feeling that homophobic and transphobic hate speeches are becoming more common in our current political state, we need to show trend so the government will take us seriously.

What is the problem and its effects?

It could make us look not as good as we can.

Current state

The current state is that we don’t have that data

Proposal for solution

A social network scraper that looks for a specific users hate speeches

Assumptions

Maavarim are not going to use it often, but when they do it will be significant

Constraints

Maavarim should pay as less as possible for 3rd parties - so free tools are prefered

OUT OF SCOPE

Other transphobic profiles detectors (not those that Maavarim are already tracking)

Metrics for success

Usable through a public link any time needed

Saved the day at least once on a Knesset meeting

Technical requirements

→ Dashboard (web interface):

Should be secured and grant access only to Maavarim employees

Displays the users that we are tracking

Ability to add more /remove users profiles or groups to track

→ Front end (web interface):

Infographic of hate speeches mount over time (2y)

→ Server end:

Scrape users profiles and groups via API key for (any of the above will be great):

Facebook
Twitter

Analyze those users public published content and use AI to determine whether the content is transphobic

Users profiles and groups should be saved in DB, it’s should be available only for the secured Maavarim dashboard

→ A database