A dataset of interactions and emotions for website user experience evaluation

Area of Science:

• Affective Computing
• Human-Computer Interaction (HCI)
• Data Science

Background:

• Understanding user emotions is crucial for improving digital experiences.
• Existing methods for emotion detection often rely on explicit user feedback.
• There is a need for datasets that link implicit user behavior to emotional states.

Purpose of the Study:

• To introduce a novel dataset for affective computing research.
• To facilitate the development of AI systems for emotion detection based on interaction data.
• To enable nuanced analysis of the interplay between user emotions and website interactions.

Main Methods:

• Collected 30 days of user interaction logs (mouse movements, key presses) from various websites.
• Paired interaction data with corresponding emotional data encoded using Ekman's emotion model.
• Structured the data for analysis in human-computer interaction and affective computing.

Main Results:

• The dataset provides a rich resource for studying emotional responses during web navigation.
• Enables correlation analysis between specific interaction patterns and emotional states.
• Offers a foundation for training AI models to infer emotions from implicit user behavior.

Conclusions:

• The dataset supports the advancement of AI-driven emotion recognition in HCI.
• It can lead to more adaptive and personalized user experiences.
• Facilitates research into how emotions influence user behavior on websites.