Case Study: RelationalAI builds a richer product knowledge graph with Diffbot

RelationalAI achieves 99.6% precision with Diffbot

The customer, RelationalAI, an AI startup, was engaged by a large online retailer to improve product recommendations but faced a challenge. Their project was hindered because the retailer's own product data was sparse and lacked the detailed attributes needed for their AI system to make accurate predictions.

Using Diffbot's automatic Product extraction technology and Knowledge Graph, RelationalAI augmented the sparse product data with information mined from the entire public web. This solution provided a complete and accurate product knowledge graph, which helped RelationalAI land the project, accelerate its delivery, and achieve a 99.6% precision rate. Furthermore, the enriched data contributed to a research paper that was accepted and published at the top-tier KDD 2020 conference.

RelationalAI