Contextual TV (CTV) dataset

Contextual TV (CTV) dataset, developed by Bang & Olufsen A/S, is a collection of 7201 self-reported TV consumption records, each enriched with detailed contextual information such as viewing situations and user attention levels. The dataset was gathered through the Experience-Sampling Method, where participants reported their TV watching habits multiple times daily for five weeks. This dataset is designed to facilitate the analysis of how genre choices correlate with various contextual factors, including the number of viewers and their attention levels. The CTV dataset aims to support the development of context-aware recommender systems, specifically focusing on understanding and predicting user preferences in dynamic viewing situations.

情境电视（Contextual TV, CTV）数据集由Bang & Olufsen A/S开发，是7201条自我报告的电视收视记录的集合，每条记录均附带详细的情境信息，例如收视场景与用户注意力水平。该数据集通过经验抽样法（Experience-Sampling Method）采集，参与者在五周周期内每日多次报告自身的电视收视习惯。本数据集旨在推动分析节目类型选择与各类情境因素（包括收视人数、用户注意力水平）之间的关联。情境电视数据集旨在支持情境感知推荐系统的开发，尤其聚焦于理解与预测动态收视场景下的用户偏好。