Empirically Testing Deep and Shallow Ranking Models for Click-Through Rate (CTR) prediction

Yi Che Yang, Ping Ching Lai, Hung Hsuan Chen

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Recent studies have reported that deep learning models perform excellently for reranking the top recommendation items. However, we found that it is not easy to reproduce some of these results. In particular, we found that recommendations based on a simple neighbor-based model, on average, outperform the results generated by deep learning models based on two datasets from e-commerce websites (one open dataset and one private dataset from our collaborating partner). Moreover, we performed an error analysis to investigate when the deep learning models perform better than simple models and when they do not. Our analysis is especially useful for medium-and small-sized online retailers that may have a smaller training dataset.

Original languageEnglish
Title of host publicationProceedings - 25th International Conference on Technologies and Applications of Artificial Intelligence, TAAI 2020
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages147-152
Number of pages6
ISBN (Electronic)9781665403801
DOIs
StatePublished - Dec 2020
Event25th International Conference on Technologies and Applications of Artificial Intelligence, TAAI 2020 - Taipei, Taiwan
Duration: 3 Dec 20205 Dec 2020

Publication series

NameProceedings - 25th International Conference on Technologies and Applications of Artificial Intelligence, TAAI 2020

Conference

Conference25th International Conference on Technologies and Applications of Artificial Intelligence, TAAI 2020
Country/TerritoryTaiwan
CityTaipei
Period3/12/205/12/20

Keywords

  • deep learning
  • match and rank
  • ranking models
  • recommender systems
  • word embedding

Fingerprint

Dive into the research topics of 'Empirically Testing Deep and Shallow Ranking Models for Click-Through Rate (CTR) prediction'. Together they form a unique fingerprint.

Cite this