Detecting Fraudulent Advertisements on a Large E-Commerce Platform
نویسندگان
چکیده
E-commerce platforms face the challenge of efficiently and accurately detecting fraudulent activity every day. Manually checking every advertisement for fraud does not scale and is financially unviable. By using automated learning algorithms, we can drastically reduce the number of advertisements that need to be checked by humans. In this paper, we present the results of a joint project with a large ecommerce company selling used goods. Using our partner’s advertisement data, we implemented several classification approaches to automatically recognize fraudulent activity. With the help of the proposed fraud detection, customer service agents only need to check about 8% of all advertisements manually for fraud. Simultaneously, we detect more than 93% of all fraudulent advertisements.
منابع مشابه
BIRDNEST: Bayesian Inference for Ratings-Fraud Detection
Review fraud is a pervasive problem in online commerce, in which fraudulent sellers write or purchase fake reviews to manipulate perception of their products and services. Fake reviews are often detected based on several signs, including 1) they occur in short bursts of time; 2) fraudulent user accounts have skewed rating distributions. However, these may both be true in any given dataset. Henc...
متن کاملData Mining Prototype for Detecting E-Commerce Fraud
The advent of electronic commerce (e-commerce) has marked a significant change in the way business now approach the implementation of their sales and marketing strategies. Electronic commerce growth has been accompanied by an increase in fraudulent practices. Researchers have proposed rules-based auditing systems for electronic commerce transactions, but they highly depend on the auditor’s know...
متن کاملIRHM: Inclusive Review Helpfulness Model for Review Helpfulness Prediction in E-commerce Platform
Online reviews have become essential aspect in E-commerce platforms due to its role for assisting customers’ buying choices. Furthermore, the most helpful reviews that have some attributes are support customers buying decision; therefore, there is needs for investigating what are the attributes that increase the Review Helpfulness (RH). This research paper proposed novel model called inclusive ...
متن کاملRating Fraud Detection - Towards Designing a Trustworthy Reputation Systems
Reputation systems could help consumers avoid transaction risk by providing historical consumers’ feedback. But, traditional reputation systems are vulnerable to the rating manipulation. It will undermine the trustworthiness of the reputation systems and users’ satisfaction will be lost. To address the issue, this study uses the real-world rating data from two travel website: Tripadvisor.com an...
متن کاملA Web-Enabled Mobile Agent Platform for E-Commerce and M-Commerce: "The Havana Platform"
A side effect to our increasingly information-driven economy and lifestyle is the annoyance and headaches of information overload. Everywhere we go, we are bombarded by email, spam, online advertisements, beepers beeping, cell phones ringing, and incoming SMS messages. The wealth of information available at our fingertips online is both a blessing and a curse in disguise. This paper presents th...
متن کامل