TY - JOUR
T1 - Angels or demons? Classifying desirable heavy users and undesirable power sellers in online C2C marketplace
AU - Yamamoto, Hikaru
AU - Sugiyama, Nina
AU - Toriumi, Fujio
AU - Kashida, Hikaru
AU - Yamaguchi, Takuma
N1 - Publisher Copyright:
© 2019, The Author(s).
PY - 2019/7/1
Y1 - 2019/7/1
N2 - To grow and succeed, online consumer-to-consumer (C2C) marketplaces need to increase the number of users and transactions, because their main revenue is usually the transaction fee. To increase the number of users and transactions, uncertainty must be reduced and a safe and enjoyable transaction environment must be maintained. In this paper, we aim to detect malicious users and power sellers who can harm the healthy growth of an online C2C platform. Using the data set of a major online C2C marketplace called Mercari, we classified undesirable users by building a classification model for banned users. The results of the banned user prediction indicated that most banned users are heavy sellers. Heavy sellers are desirable from the viewpoint of increasing the transaction fee revenue, but many are power sellers who are running full-time businesses on the platform, making it difficult for non-professional sellers to compete, and their dominance may eventually alienate users. Thus, we built another classification model to classify desirable and undesirable power sellers. Applying the model to the CART classifier, we successfully classified non-professional heavy users and undesirable power sellers in an online C2C marketplace.
AB - To grow and succeed, online consumer-to-consumer (C2C) marketplaces need to increase the number of users and transactions, because their main revenue is usually the transaction fee. To increase the number of users and transactions, uncertainty must be reduced and a safe and enjoyable transaction environment must be maintained. In this paper, we aim to detect malicious users and power sellers who can harm the healthy growth of an online C2C platform. Using the data set of a major online C2C marketplace called Mercari, we classified undesirable users by building a classification model for banned users. The results of the banned user prediction indicated that most banned users are heavy sellers. Heavy sellers are desirable from the viewpoint of increasing the transaction fee revenue, but many are power sellers who are running full-time businesses on the platform, making it difficult for non-professional sellers to compete, and their dominance may eventually alienate users. Thus, we built another classification model to classify desirable and undesirable power sellers. Applying the model to the CART classifier, we successfully classified non-professional heavy users and undesirable power sellers in an online C2C marketplace.
KW - Banned user prediction
KW - Consumer-to-consumer marketplace
KW - Online fraud
KW - Online platform
UR - http://www.scopus.com/inward/record.url?scp=85096060510&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85096060510&partnerID=8YFLogxK
U2 - 10.1007/s42001-019-00050-y
DO - 10.1007/s42001-019-00050-y
M3 - Article
AN - SCOPUS:85096060510
VL - 2
SP - 315
EP - 329
JO - Journal of Computational Social Science
JF - Journal of Computational Social Science
SN - 2432-2717
IS - 2
ER -