Slopk

Slopk, as known as Slop Killer, is a toolkit designed to combat low-quality and inappropriate content on social media platforms, such as TikTok. The toolkit comprises of a statistics app and an API.

Comment

Inspiration

TikTok Developer API, https://arxiv.org/html/2505.10867v1 Youtube Analytics, SocialBlade

What it does

SlopK Statistics & Finder toolkit is designed to help users, creators and administrators alike, to discern the 'good' content creators from the 'poor' ones AKA slop content. By utilizing AssemblyAI API, PyTube and Lynx, SlopK Statistics, paired with Finder is able to summarise the content from videos, allowing for the end-users to gauge the quality and type of content more quickly and efficiently. SlopK Finder is also able to find any sensitive topics discussed in the video. SlopK Finder is able to identify the topics that are being discussed in the video along with its confidence score for each category. This will give users a trigger warning and might even help with age restriction.

BUSINESS USE CASES:

  1. Users will be able to discern good content from slop content by reading the summary of the video generated by AssemblyAI API. The video containing good content will most probably produce a concise, yet insightful summary while the slop content's summary will be about some fake AI generated love story about some teenagers(yucks).
  2. Flag out and filter inappropriate content.
  3. Accounts that produce content with similar topics on a frequent basis can potentially be guilty of producing slop content.

Overall, SlopK Finder intends to improve and encourage the development of Value Sharing capabilities on these platforms.

How we built it

  • react
  • lynx
  • streamlit
  • npm packages in requirements.txt
  • tailwind

Challenges we ran into

Finding models that will achieve our goal, and creating a responsive UI/UX application.

Accomplishments that we're proud of

  1. With little knowledge, developing an API that can do language parsing and summaries of videos.
  2. Using Lynx to develop an application within 8 hours.

What we learned

Developing in Python, Lynx, video pitching and coming up with out of the box ideas.

What's next for Slopk

Advanced models and an integrated application to manage statistics and videos (hopefully on TikTok).

Built With

Share this project:

Updates