Predicting CTR of Responsive Search Ads Using Handcrafted Features

Melvin Charles Ortua DY

doi:10.11517/pjsai.JSAI2023.0_1U5IS2b05

37th (2023)

Session ID : 1U5-IS-2b-05

DOI https://doi.org/10.11517/pjsai.JSAI2023.0_1U5IS2b05

Conference information

Host: The Japanese Society for Artificial Intelligence

Name : The 37th Annual Conference of the Japanese Society for Artificial Intelligence

Number : 37

Location : [in Japanese]

Date : June 06, 2023 - June 09, 2023

Predicting CTR of Responsive Search Ads Using Handcrafted Features

*Melvin Charles Ortua DY

Author information

Keywords: Machine Learning, Feature Engineering, Advertising

CONFERENCE PROCEEDINGS FREE ACCESS

Details

Abstract

In this paper, I demonstrate that a reasonably sized set of handcrafted features (866, applied to titles and description texts separately) plus encoded metadata can be used to predict the click-through rates of the dynamic Responsive Search Ad format, exceeding the performance of some fine-tuned Transformer-based large language models at a fraction of the training cost.

Corresponding author

Conference information

Register with J-STAGE for free!