Sunandan Chakraborty is a Moore-Sloan Postdoctoral Researcher at CDS and a visiting scholar at the Environmental Studies Department [https://environment.as.nyu.
Can you give us a bit of background on yourself? How did you get to where you are now?
When I was completing my Master’s degree at the Indian Institute of Technology Kharagpur (IIT) and working for Microsoft Research India, I realized that I wanted to develop my interests in Information Extraction, Natural Language Processing, and Machine Learning. I applied to NYU because of its strong research groups in my areas of study.
Why did you choose to examine the subject of illegal wildlife trading?
Illegal wildlife trade on the web is a major problem that is often overlooked, and the proposals that do address this problem remain inadequate. I felt that a good solution was urgently needed, especially for the law enforcement agencies. The problem’s computational complexity was also an intriguing challenge for me, as issues on such a large scale often require particularly innovative data science solutions. Additionally, I have always been committed to using computational and data-driven methods to conduct research on real world problems in developing regions.
What sort of information or trends did you go in looking for?
The main trends I interrogated were as follows. Firstly, what are the main species and products involved in wildlife trade? Secondly, what are the main sites that are used as a trading platform for these illegal products? Finally, what are the major hot spots for this activity?
At any point in your research did your objectives change? Were there any assumptions you had going in that had to be adjusted once you started to analyze the data?
Our objective did not change, but we did modify our approach to account for some of our assumptions. This involved changing aspects like the list of keywords, adding more sites to the list which had not seemed to be important before, and adding more common names and code words for certain species.
With the tool that you designed, are you looking for words, code-words, pictures, something else, or a combination?
It is a combination of the things mentioned above. There are texts associated with the online advertisements and postings, so they contain specific product information like its price, shipping formation, and item origin. Recording these details were crucial to our problem-solving process. We are presently analyzing the images in these ads by designing an item/species identification tool using the state-of-the-art computer vision techniques.
How much supervision is required from humans when using your tool?
The tool is being designed in such a way that the end users will need minimum supervision. During the developmental phase, however, a certain amount of supervision is needed, particularly from industry experts. We rely on their expertise to identify illegal items in these advertisements, and which properties are used to detect illegal objects/species.
Where are you predominantly looking for illegal wildlife trading? eBay? The Dark Web?
Illegal wildlife trade is flourishing in the open web: we have so far identified certain online marketplaces and retail sites across different countries. Interestingly, a recent study found no evidence of illegal wildlife trade in the dark web.
How has your time with the CDS program been so far?
Very enjoyable. It is a very lively and vibrant community; it is a pleasure to work within this environment. I enjoy being in this interdisciplinary environment. Apart from my main project, I have had the chance to collaborate with other researchers from different backgrounds, and work on several interdisciplinary projects. This gives me the opportunity to learn skills from different communities as well as contribute in research domains other than my own.
by Cherrie Kwok