<abstract xmlns="http://www.w3.org/1999/xhtml"><p>This paper presents a proposition to utilize flexible neural network architecture called Deep Hybrid Collaborative Filtering with Content (DHCF) as a product recommendation engine. Its main goal is to provide better shopping suggestions for customers on the e-commerce platform. The system was tested on 2018 Amazon Reviews Dataset, using repeated cross validation and compared with other approaches: collaborative filtering (CF) and deep collaborative filtering (DCF) in terms of mean squared error (MSE), mean absolute error (MAE) and mean absolute percentage error (MAPE). DCF and DHCF were proved to be significantly better than the CF. DHCF proved to be better than DCF in terms of MAE and MAPE, it also scored the best on separate test data. The significance of the differences was checked by means of a Friedman test, followed by post-hoc comparisons to control p-value. The experiment shows that DHCF can outperform other approaches considered in the study, with more robust scores.</p></abstract>

This paper presents a proposition to utilize flexible neural network architecture called Deep Hybrid Collaborative Filtering with Content (DHCF) as a product recommendation engine. Its main goal is to provide better shopping suggestions for customers on the e-commerce platform. The system was tested on 2018 Amazon Reviews Dataset, using repeated cross validation and compared with other approaches: collaborative filtering (CF) and deep collaborative filtering (DCF) in terms of mean squared error (MSE), mean absolute error (MAE) and mean absolute percentage error (MAPE). DCF and DHCF were proved to be significantly better than the CF. DHCF proved to be better than DCF in terms of MAE and MAPE, it also scored the best on separate test data. The significance of the differences was checked by means of a Friedman test, followed by post-hoc comparisons to control p-value. The experiment shows that DHCF can outperform other approaches considered in the study, with more robust scores.

Improvement of E-Commerce Recommendation Systems with Deep Hybrid Collaborative Filtering with Content: A Case Study

Wroclaw University of Economics and Business

Econometrics

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 3.0 License.

This paper presents a proposition to utilize flexible neural network architecture called Deep Hybrid Collaborative Filtering with Content (DHCF) as a product...

Improvement of E-Commerce Recommendation Systems with Deep Hybrid Collaborative Filtering with Content: A Case Study

Data publikacji: 24 lis 2020

Zakres stron: 37 - 50

DOI: https://doi.org/10.15611/eada.2020.3.03

Słowa kluczowe
collaborative filtering, deep learning, content model, product recommendation

© 2020 Filip Wójcik et al., published by Sciendo

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 3.0 License.

Improvement of E-Commerce Recommendation Systems with Deep Hybrid Collaborative Filtering with Content: A Case Study

Data publikacji: 24 lis 2020

Zakres stron: 37 - 50

DOI: https://doi.org/10.15611/eada.2020.3.03

Słowa kluczowecollaborative filtering, deep learning, content model, product recommendation

© 2020 Filip Wójcik et al., published by Sciendo

This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 3.0 License.

Słowa kluczowe
collaborative filtering, deep learning, content model, product recommendation