Dataset Overview

This dataset is derived from Terms of Service; Didn't Read (ToS;DR), a project that analyzes and categorizes terms of service from various online services. The dataset has been cleaned and organized into two CSV files, with a focus on reproducibility and usability. The privacy dataset is a subset of the full dataset, specifically filtering for privacy-related terms.

File Descriptions

1. training_tosdr_all_data.csv

This file contains the complete collection of terms of service data after cleaning and preprocessing. Each row represents a statement (or "point") extracted from a service's terms of service.

Key Columns:

• case_id: Unique identifier for the case.
• case_title: Brief description of the case.
• topic_id: Unique identifier for the topic.
• topic_title: Broad category the case falls under (e.g., Transparency, Copyright License).
• sentence: The extracted text from the terms of service.
• seq_case_id: Sequential identifier for the case, used for mapping.
• seq_topic_id: Sequential identifier for the topic, used for mapping.

2. training_tosdr_privacy_data.csv

This file is a subset of the full dataset, focusing exclusively on privacy-related terms. It includes cases related to tracking, data collection, account deletion policies, and other privacy-related topics.

Key Columns:

• case_id: Unique identifier for the case.
• case_title: Brief description of the case.
• topic_id: Unique identifier for the topic.
• topic_title: Broad category the case falls under (e.g., Privacy, Data Collection).
• sentence: The extracted text from the terms of service.
• seq_case_id: Sequential identifier for the case, used for mapping.
• seq_topic_id: Sequential identifier for the topic, used for mapping.