본문 바로가기
  • Home

Data-Based Contest Analysis Framework Design: Using Text Mining to Derive Insights for Contest Planning

  • Journal of Internet of Things and Convergence
  • Abbr : JKIOTS
  • 2025, 11(3), pp.11~17
  • Publisher : The Korea Internet of Things Society
  • Research Area : Engineering > Computer Science > Internet Information Processing
  • Received : April 3, 2025
  • Accepted : May 13, 2025
  • Published : June 30, 2025

Chae-Jin LIM 1 Jung Soo Han 2 Hyun-Seob Lee 2

1백석대학교 소프트웨어융합전공
2백석대학교

Accredited

ABSTRACT

In 2014, the number of contests held in Korea was about 4,000, and it is expected to increase to about 10,000 by 2024. The number of contests is steadily increasing by sector. It's becoming increasingly important for contest organisers to select topics efficiently to maximise impact while reducing costs. This increases the need for data-driven analytical tools to help them do this. In this study, we propose a framework to analyse 7,587 (5,646 after deduplication) comic contest data from 2011 to 2024 using text mining techniques. Through an analysis methodology that combines TF-IDF and Word2Vec, we propose a methodology that reflects the frequency of keyword occurrence and semantic associations together, and comprehensively identifies trends over time to provide practical insights for contest planners. The framework creates a ranking of important keywords through a composite vector of frequency-based importance scores and semantics. Finally, it calculates a trend metric that spans the entire time period by averaging the percentage change from year to year.

Citation status

* References for papers published after 2023 are currently being built.