Conversion Prediction in Google Search Ads Keyword Selection Using the K-Nearest Neighbor and C4.5 Algorithms

Abstract

This study was conducted to analyze and compare the effectiveness of two algorithms—K-Nearest Neighbor (K-NN) and C4.5—in predicting keyword conversion on the Google Ads platform. With the rapid growth of digital marketing, selecting the right keywords has become crucial for improving conversion rates. The research utilized a dataset of 673 entries with 12 relevant attributes, collected from historical ads and the Google Ads Keyword Planner. A comparative experimental approach was employed, with the data split into training (80%) and testing (20%) sets. The analysis revealed that the C4.5 algorithm achieved higher accuracy (85.41%) compared to K-NN (74.86%). Evaluation was based on metrics such as accuracy, precision, recall, and F1-score, which indicated that C4.5 was more effective in predicting conversions using the given dataset. These findings offer valuable insights for advertisers aiming to optimize their ad campaigns by selecting more effective keywords. However, the study also acknowledges limitations and recommends further research using larger and more diverse datasets to enhance model accuracy.