ViTOR: Learning to Rank Webpages Based on Visual Features

We introduce the Visual learning TO Rank (ViTOR) model that integrates state-of-the-art visual features extraction methods: (i) transfer learning from a pre-trained image classification model, and (ii) synthetic saliency heatmaps generated from webpage snapshots. Since there is currently no public dataset available for the task of LTR with visual features, we also introduce and release the ViTOR dataset, containing visually rich and diverse webpages.

The ViTOR dataset consists of visual snapshots, non-visual features and relevance judgments for ClueWeb12 webpages and TREC Web Track queries. We experiment with the proposed ViTOR model on the newly introduced ViTOR dataset and show that our model significantly improves the performance of LTR with visual features.