Journal
ELECTRONIC COMMERCE RESEARCH AND APPLICATIONS
Volume 50, Issue -, Pages -Publisher
ELSEVIER
DOI: 10.1016/j.elerap.2021.101107
Keywords
Click farming; Online goods; PU learning; Weighted random forest; Taobao
Categories
Funding
- National Natural Science Foundation of China [71671056, 91846201, 71801108, 72171071]
- Humanity and Social Science Foundation of the Ministry of Education of China [19YJA790035]
- National Statistical Science Research Projects of China [2019LD05]
Ask authors/readers for more resources
The study analyzed and compared the performance of click farming across different categories of online goods. New features were extracted and investigated for their effectiveness in identifying click farming, with results showing that click farming is most common in clothing-related goods.
Click farming is common in online shopping. It is thus important to identify click farming and compare its performance across different categories of online goods. To this end, we conduct an empirical analysis of click farming on the Taobao platform in China. First, we extract several new features from three sources, namely main goods, online shop itself, and online reviews, based on the formation mechanism of click farming. Second, we investigate their usefulness in identifying click farming among different online goods, including importance analysis and partial dependence analysis. Third, we further investigate the contribution of constructed features to predicting click farming. Our findings confirm the effectiveness of our created features and the heterogeneity of click farming among different online goods. Specifically, click farming is most likely to happen in clothingrelated goods, then followed by electronic goods and service-related goods. Our results are significant for consumers to understand online information and for online business platforms to reduce the occurrence of click farming.
Authors
I am an author on this paper
Click your name to claim this paper and add it to your profile.
Reviews
Recommended
No Data Available