ThreatLens

Inspiration

Our goal is to create an online harassment website to detect threats and/or harassment. With the era of social media, online harassment has been ever so prevailent. Forensic applications, such as detection of sentimental analysis, have been more applicible in solving crimes. Thus, ThreatLens© was created. This website can be used for forensic analysis to detect online users that may pose as a threat.

What it does

ThreatLens© uses the consumer's Tiktok username to obtain comments in their Tiktok videos. It scraps Tiktok comments and detects texts that are threats and/or harassments. Natural Language Processing (NLP) is used for sentimental analysis. Data cleaning was done to remove stopwords. A ranked list is then created to showcase users that are potientially harassers based on their comments.

How we built it

The frontend of the website is created with React, and the backend is created with Python and Django. Json is used for collecting Tiktok's comments.

Challenges we ran into

With only 48 hours, time was limited. Our main challenges were to connect the backend sentimental analysis (Python) to Django. Another main challenge was creating a sentimental analysis that includes detection of harmful emojis and words.

Accomplishments that we're proud of

Completing this project and not giving up! :)

What we learned

We learned React, strengthened our Python skills, scraping comments off of Tiktok, and collaborating as a team.

What's next for ThreatLens

We're hoping to improve the matching algorithm of the user's personal information (i.e., their occupation and ethnicity) and target comments that may be related to their basic information to harass the user. This will improve the forensic capability of detecting threats and harassments via sentimental analysis.