May loko sa kanto: Uncovering user personas matched to possible criminal behavior in Locanto PH listings
Document Type
Article
Source Title
AIM Webinar Data Mining And Wrangling Public Presentation
Publication Date
8-2020
Abstract
In this pandemic that we are experiencing right now, it is inevitable that us Filipinos, will do even more of our activities online given the already significant immersion we have in this digital world. Included in these set of activities are personal ones that address our mental and emotional needs. One of the mediums that people can do this is through posting in online classified ads, indicating maybe some personal details that one may be willing to share to the public. This causes one to be vulnerable or at least exposed to some kinds of illicit activity including scams, and even signs of prostitution done also by other people. Locanto Philippines is one of these online classified ads websites that enable these people at both ends of the spectrum to freely conduct these activities. By analyzing the listing descriptions in Term Frequency – Inverse Document Frequency format, the group was able to uncover topics and themes including monetary transactions that are incorporated with sexual activities given words or phrases including money, pay, and willing to pay. Through Agglomerative clustering using Ward’s distance, the group was able to determine five (5) clusters of the listings, two of which are those that intend to look for long-term or serious relationships and those which intend to continue the transaction externally (outside of the website), usually though Viber, WhatsApp, or personal contact. The former was found out to be associated with the themes determined earlier regarding sex and money, and to online dating scams as well according to external articles. The latter was deemed suspicious as well due to imposing further discussions externally which may be normal, but these seem to be associated with the said sex for money activities as well. These study was able to determine at least the solicitation of prostitution and possible leads for checking with a bigger scaled data with more coverage and more substantial causes for suspicion.
html
Recommended Citation
Delgado, K. P. (2020). May loko sa kanto: Uncovering user personas matched to possible criminal behavior in Locanto PH listings. AIM Webinar Data Mining And Wrangling Public Presentation Retrieved from https://animorepository.dlsu.edu.ph/faculty_research/11129
Disciplines
Advertising and Promotion Management | Computer Sciences
Keywords
Personals; Internet advertising; Internet fraud
Upload File
wf_no
Note
Title taken from the file namen and transmission letter