Webis Generated Native Ads 2025

Dataset Summary This dataset was created to train ad blocking systems on the task of identifying advertisements in the responses of large language models (LLMs) and search engines that use retrieval-augmented generation (RAG). It is the successor of the Webis Generated Native Ads 2024 dataset. Citation @misc{heineking:2025, author = {Sebastian Heineking and Ines Zelch and Wilhelm Pertsch and Christian Deubel and Matthias Hagen and Martin Potthast}, title = {{Webis Generated Native Ads 2025}}, doi = {10.5281/zenodo.16941607}, year = 2025}## 数据集摘要 本数据集专为训练广告拦截系统构建，旨在完成识别使用检索增强生成（Retrieval-Augmented Generation，RAG）技术的大语言模型（Large Language Model，LLM）与搜索引擎返回结果中的广告这一任务。本数据集是Webis Generated Native Ads 2024数据集的后继版本。 ## 引用 @misc{heineking:2025, author = {Sebastian Heineking and Ines Zelch and Wilhelm Pertsch and Christian Deubel and Matthias Hagen and Martin Potthast}, title = {{Webis 生成原生广告 2025}}, doi = {10.5281/zenodo.16941607}, year = 2025 }