Discord Bot for Video Game Recommendations Using Collaborative Filtering and TF IDF Methods
bit-Tech · 10.08.2026 · 9 (1) · 801-812
DOI: 10.32877/bt.v9i1.3983 ↗Kurz gefasst
Fachartikel in der Zeitschrift bit-Tech, online erschienen am 10.08.2026. Verfasst von Dewa Aryana und 2 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY-SA 4.0).
Abstract (Originalsprache)
Game selection in Discord communities often involves fragmented and subjective suggestions, creating information overload for gamers seeking titles that match their preferences. This study aims to develop and evaluate a Discord-based video game recommendation bot that integrates TF-IDF-based content representation with Collaborative Filtering to generate ranked game recommendations within a chat-based community environment. The system used two secondary Kaggle datasets: 2,390 Steam game metadata records containing game names, genres, categories, labels, and descriptions, and 70,490 user-behavior records containing user IDs, owned games, and total playtime. Textual metadata were cleaned, case-folded, tokenized, filtered through stopword removal, lemmatized, and transformed into TF-IDF vectors, with semantic representation supported through a 90% TF-IDF and 10% SBERT weighting scheme. Collaborative Filtering was then applied to user gameplay behavior, and the final hybrid ranking combined content seed score, TF-IDF similarity, and collaborative similarity before being deployed through a Discord bot interface. Evaluation was conducted on 56 test samples from a reduced dataset of 278 records using an 80:20 split. The model achieved HitRate@5 of 96.36%, Recall@5 of 96.36%, and Precision@5 of 64.73%, indicating strong retrieval of at least one relevant game in the top-five results but only moderate ranking precision. These findings show that the proposed hybrid bot is promising for Discord-based game discovery, while broader validation with larger datasets, baseline comparisons, and real-user testing remains necessary.
Zitieren
APA
Aryana, D., Rahmat, B., & Haromainy, M. M. A. (2026). Discord Bot for Video Game Recommendations Using Collaborative Filtering and TF IDF Methods. bit-Tech, 9(1), 801-812. https://doi.org/10.32877/bt.v9i1.3983
Vancouver
Aryana D, Rahmat B, Haromainy MMA. Discord Bot for Video Game Recommendations Using Collaborative Filtering and TF IDF Methods. bit-Tech. 2026;9(1):801-812. doi:10.32877/bt.v9i1.3983
BibTeX
@article{Aryana2026,
author = {Aryana, Dewa and Rahmat, Basuki and Haromainy, Muhammad Muharrom Al},
title = {Discord Bot for Video Game Recommendations Using Collaborative Filtering and TF IDF Methods},
journal = {bit-Tech},
year = {2026},
volume = {9},
number = {1},
pages = {801--812},
doi = {10.32877/bt.v9i1.3983}
}
EUVIMED