Leveraging LLMs and Graph-based representation for Keyword Ads Optimization

Keyword Ads, one of the most popular types of digital advertising, involve using targeted words to display ads based on user search queries. While Large Language Models (LLMs) can enhance this process by understanding search queries and generating relevant keywords, they may struggle with understanding complex relationships between products and keywords relevant to a given business or platform. That is where Graph-based representations can come in handy and address these shortcomings and help improve the relevance of the ads, leading to better customer experience and ROI(Return of Investment) for the ads partners.

In this session, we will understand various components of keywords based ads and how LLMs and Graph based representation can be leveraged to improve the performance of these components.