Ecosyste.ms: Awesome

An open API service indexing awesome lists of open source software.

Awesome Lists | Featured Topics | Projects

https://github.com/edochiari/tiktok-project

This project builds a predictive model to help TikTok classify user-reported content claims, improving moderation efficiency by identifying and prioritizing content that may need review. Insights from this model enable TikTok to manage reports more effectively, ensuring a safer and more engaging platform.
https://github.com/edochiari/tiktok-project

content-claims dataanalysis datacleaning hypothesis-testing jupyter-notebook regression tiktok

Last synced: about 17 hours ago
JSON representation

This project builds a predictive model to help TikTok classify user-reported content claims, improving moderation efficiency by identifying and prioritizing content that may need review. Insights from this model enable TikTok to manage reports more effectively, ensuring a safer and more engaging platform.

Awesome Lists containing this project

README 

        
# TikTok Claims Classification Project



## Overview

This project focuses on developing a predictive model to support TikTok’s moderation team by classifying user-submitted content claims efficiently. By analyzing user reports on videos and comments, the project aims to build a model that distinguishes between content with user claims versus opinions. This approach will help TikTok reduce the backlog of reports, prioritize moderation efforts, and maintain a safe and engaging community.



## Project Goals

1. **Classify Content Claims**: Build and evaluate a model to predict whether a video contains a claim or an opinion, enabling TikTok to streamline content moderation.

2. **Enhance Moderation Efficiency**: Provide a scalable solution for handling high volumes of user reports, improving the speed and accuracy of content review processes.

3. **Deliver Insights for Stakeholders**: Generate actionable insights from user reports to aid TikTok leadership in understanding content trends and moderation needs.



## Deliverables

The final project deliverables include:



- **Model Evaluation**: Comprehensive assessment of the classification model, including accuracy, precision, and recall, to gauge its effectiveness in content claim prediction.

- **Data Visualizations**: Interactive Tableau dashboards summarizing user report trends, claim types, and other key insights, accessible to non-technical stakeholders.

- **Feature Analysis**: Examination of features that contribute most to accurate claim classification, with discussions on potential causative relationships.

- **Future Model Improvements**: Recommendations for additional features and data sources that may enhance the accuracy and relevance of the model.



## Tools and Libraries Used

- **Data Analysis and Visualization**: Pandas, NumPy, Matplotlib, Seaborn, Tableau

- **Machine Learning**: Scikit-learn (for regression and classification models)

- **Notebook Environment**: Jupyter Notebook



## Project Structure

The project is organized as follows:



1. **Data Preparation**: Building and organizing a comprehensive dataset from user reports for claims classification, ensuring data quality and suitability for analysis.

2. **Exploratory Data Analysis (EDA)**: Analyzing user reports to identify claim patterns, trends, and factors that may impact claim classification.

3. **Hypothesis Testing**: Conducting hypothesis tests to determine the significance of various factors within user reports, informing model feature selection.

4. **Model Building and Evaluation**: Developing and testing a regression model to classify content claims, followed by evaluation using key performance metrics.

5. **Executive Summary**: A presentation-ready summary for stakeholders, highlighting findings, model performance, and potential impact on moderation efforts.



## Conclusion

This project offers TikTok a data-driven approach to improve content moderation by predicting user claims more effectively. With a model that enhances the prioritization of user reports, TikTok can maintain a safe, enjoyable platform while efficiently managing moderation resources.



## Badges



Add badges from somewhere like: [shields.io](https://shields.io/)



[![MIT License](https://img.shields.io/badge/License-MIT-green.svg)](https://choosealicense.com/licenses/mit/)

[![GPLv3 License](https://img.shields.io/badge/License-GPL%20v3-yellow.svg)](https://opensource.org/licenses/)

[![AGPL License](https://img.shields.io/badge/license-AGPL-blue.svg)](http://www.gnu.org/licenses/agpl-3.0)