Development of e-Procurement Datasets Based on e-Katalog 5.0 Services in Indonesia

The recommender system (RS) has been successful in recommending products needed by customers. RS has been implemented in many domains such as entertainment, news, e-commerce and social media. However, RS research in the domain of e-Procurement in the area of e-Government is still very limited. There...

Full description

Saved in:
Bibliographic Details
Published in2024 12th International Conference on Information and Communication Technology (ICoICT) pp. 521 - 527
Main Authors Fajar Surya Gumilang, Soni, Adiwijaya, Baizal, Z. K. A., Toto Wibowo, Agung, Michael Fritz, James
Format Conference Proceeding
LanguageEnglish
Published IEEE 07.08.2024
Subjects
Collaborative Filtering
Companies
Dataset
e-Procurement
Electronic government
Entertainment industry
Filtering
Information and communication technology
Procurement
Qualifications
Recommender System
Recommender systems
Social networking (online)
Testing
Online AccessGet full text

Cover

More Information
Summary:The recommender system (RS) has been successful in recommending products needed by customers. RS has been implemented in many domains such as entertainment, news, e-commerce and social media. However, RS research in the domain of e-Procurement in the area of e-Government is still very limited. Therefore, this research proposes a novel and qualified dataset for RS in the e-Procurement domain based on the Indonesian e-Katalog 5.0 service. The dataset is ready to be used using various methods in Collaborative Filtering (CF)-based RS. To produce the dataset, we carry out a data scrapping process, and data preparation by carrying out Summing and Re-scaling methods based on several existing rating criteria. In addition, we operate several algorithms in CF by utilizing our proposed dataset. For all the methods, the obtained Precision@10 and Recall@10 are above 0.9, which indicates the good performance. Apart from that, the RMSE, MAE and NDCG results also show good performance for all CF methods. The extensive experiments have demonstrated that our proposed dataset can be used properly in CF-based RS, specifically in e-Procurement domain.
DOI:10.1109/ICoICT61617.2024.10698392