Abstract—Any business that operates on the Internet and accepts payments through debit or credit cards, also implicitly accepts that some transaction may be fraudulent. The design of effective strategies to face this problem is challenging, due to factors such as the heterogeneity and the non stationary distribution of the data, as well as the presence of an imbalanced class distribution, and the scarcity of public datasets. Differently from the state-of-the-art strategies, instead of producing a unique model based on the past transactions of the users, our approach generates a set of models (behavioral patterns) to evaluate a new transaction, by considering the behavior of the user in different temporal frames of her/his history. By using only the legitimate past transactions of a user, we can operate in a proactive manner, by detecting the fraudulent ones that have never occurred. This also overcomes the data imbalance that afflicts the state-of-the-art approaches. We evaluate our proposal by comparing it with one of the most performing approaches at the state of the art (i.e., Random Forests), using a real-world credit card dataset.

Index Terms—Fraud detection, pattern mining, rule learning.

Roberto Saia, Ludovico Boratto, Salvatore Carta are with the Dipartimento di Matematica e Informatica, Università di Cagliari, Via Ospedale 72 - 09124 Cagliari, Italy.

Cite: Roberto Saia, Ludovico Boratto, Salvatore Carta, " A Proactive Time-frame Convolution Vector (TFCV) Technique to Detect Frauds Attempts in e-Commerce Transactions," International Journal of e-Education, e-Business, e-Management and e-Learning vol. 5, no. 4, pp. 229-236, 2015.