# ViTOR: Learning to Rank Webpages Based on Visual Features

## TL;DR

This paper presents ViTOR, a novel learning to rank model that leverages visual features from webpage snapshots, using transfer learning and saliency maps, and introduces a new dataset for this task.

## Contribution

The paper introduces ViTOR, a new model combining visual features with LTR, and releases the first public dataset for visual webpage ranking.

## Key findings

- ViTOR significantly improves ranking performance with visual features.
- Transfer learning enhances visual feature extraction for ranking.
- Saliency maps contribute to better webpage relevance estimation.

## Abstract

The visual appearance of a webpage carries valuable information about its quality and can be used to improve the performance of learning to rank (LTR). We introduce the Visual learning TO Rank (ViTOR) model that integrates state-of-the-art visual features extraction methods by (i) transfer learning from a pre-trained image classification model, and (ii) synthetic saliency heat maps generated from webpage snapshots. Since there is currently no public dataset 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 ViTOR dataset and show that it significantly improves the performance of LTR with visual features