Multimodal Scoring of Website Content Using Vision-Language Models for SEO Objectives

Businesses are increasingly using personal websites to introduce themselves and sell their products or services. This has intensified competition among websites to attract a larger audience. As a result, website content production aimed at improving SEO scores has become a booming market. This research seeks to measure the alignment between website content and its topic using vision-language models, ranking websites accordingly. The results indicate that the proposed ranking method enhances accuracy compared to purely NLP-based approaches.