Information Architecture: Using K-Means Clustering and the Best Merge Method for Open Card Sorting Data Analysis. Issue 6 (27th July 2022)

• Record Type: Journal Article
• Authors: Paea, Sione; Katsanos, Christos; Bulivou, Gabiriele
• Is Part Of: Interacting with computers. Volume 33:Issue 6(2021)
• Page Start: 670
• Page End: 689
• Publication Date: 2022-07-27
• Subjects: information architecture -- dendrograms -- category -- hypertext/hypermedia -- user-centered design -- HCI design and evaluation methods; Human-computer interaction -- Periodicals; 004.019
• DOI: 10.1093/iwc/iwac022
• Languages: English
• ISSNs: 0953-5438

Abstract

Abstract: Open card sorting is a well-established method for discovering how people understand and categorize information. This paper addresses the problem of quantitatively analyzing open card sorting data using the K-means algorithm. Although the K-means algorithm is effective, its results are too sensitive to initial category centers. Therefore, many approaches in the literature have focused on determining suitable initial centers. However, this is not always possible, especially when the number of categories is increased. This paper proposes an approach to improve the quality of the solution produced by the K-means for open card sort data analysis. Results show that the proposed initialization approach for K-means outperforms existing initialization methods, such as MaxMin, random initialization and K-means++. The proposed algorithm is applied to a real-world open card sorting dataset, and, unlike existing solutions in the literature, it can be used with any number of participants and cards.