Readsight is a pre-launch data-driven startup focused on improving journalism analytics through AI. Our mission is to enhance the quality of journalism and shift users to engage more meaningfully with digital articles.
Our immediate goal is to develop a model that generates retention heatmaps using just scroll position data. This experimental approach aims to predict where readers' eyes focus when leaving a webpage and which specific words they were reading at exit points, as well as finding areas of attention loss and high emotional response.
We're developing a data collection mechanism where volunteer participants undergo brief calibration before reading various articles embedded with our tracking module. We log scroll position and eye movement data to train and validate our machine learning models.
Our machine learning model processes scroll position data to create attention distribution and loss heatmaps. The prediction model integrates metrics for engagement, enjoyment, and insightfulness of article content.
Privacy matters deeply to us. The data we collect is entirely non-identifying. All we collect is scrolling data, which is then processed through our advanced machine learning models to analyze reader retention patterns.
Developing our volunteer research platform and implementing advanced tracking mechanisms to gather high-quality data. Simultaneously, we're designing our initial prediction algorithms and integrating scroll pattern analysis with eye movement tracking to create our first generation of attention heatmaps.
Fully operational data-collection site and prototype prediction model ready for initial testing and validation with publishing partners.
We're currently assembling a group of volunteers for data collection.
Volunteer Portal Coming SoonInterested in what we're building or looking to collaborate? I'd love to hear from you.
Email: [email protected]