Electronic News Dataset for Native Advertisement Detection.

Area of Science:

• Digital Marketing
• Computational Linguistics
• Information Science

Background:

• Native advertising is a prevalent online marketing strategy that integrates sponsored content seamlessly with editorial material, often deceiving readers.
• The sophisticated integration of native ads across various formats (text, video, social media) poses significant detection challenges.
• Ensuring transparency in online environments requires robust methodologies and comprehensive datasets for identifying native advertising.

Purpose of the Study:

• To develop a specialized, systematically collected, and annotated dataset for enhancing native advertising detection.
• To facilitate the creation and evaluation of advanced algorithms for identifying native ads and their implicit characteristics.
• To promote transparency in online advertising by distinguishing sponsored content from genuine editorial material.

Main Methods:

• Systematic data collection and annotation of native advertising content.
• Aggregation of news articles from six major Indonesian electronic news portals.
• Inclusion of four implicit characteristics beyond simple native ad identification.

Main Results:

• Creation of a meticulously annotated dataset specifically designed for native ad detection.
• The dataset serves as a crucial resource for developing and testing sophisticated detection algorithms.
• The research provides a foundation for improved accuracy in distinguishing native ads from editorial content.

Conclusions:

• The developed dataset significantly enhances the capability to detect native advertising.
• This resource is vital for advancing research in automated native ad identification.
• The findings contribute to greater transparency and ethical practices in online advertising ecosystems.