Kushal Shah

Digital Financial Services Research Group

---

Advised primarily by Prof. Richard Anderson to undertake econometric analysis on various datasets pertaining to use and access of financial services in emerging economies.

1. Cross-Refrencing survey data-sets by Helix Institute of Digital Finance and Intermedia - we are trying to understand user perception of fraud while using various financial services.
2. We are trying to quantify the relationship between gender and its impact on financial inclusion in Pakistan, India and Bangladesh.

Collaborators: Samia Ibtasam Razaq, Sam Castle, Fahad Pervaiz

• Date: September 2016 - Present

Close

Investigation of Phone Upgrade Behavior in Rural Areas

---

Working with Prof. Kurtis Heimerl and Facebook's data science team to understand adoption of 4G phones in rural areas. The aim is to assess the need for Facebook to upgrade their existing technology based on user behavior - assessed through Call Data Records.

• Date: Jan 2017 - Present

Close

Crisis Informatics

---

Working with Prof. Kate Starbird to understand what data is actionable data for different stakeholders involved in crisis relief. Its a qualitative project where we are undertaking interviews and surveys of humanitarian aid responders, local civic authorities, and online volunteers to observe underlying differences on what data is useful to them and how they use it.

Collaborators: Himanshu Zade, Dr. Muhammad Imran, Vaibhavi Rangarajan, Priyanka Kshirshagar, Gabby Milano

• Date: September 2016 - Present

Close

Urban Computing

---

As a Research Assistant at Qatar Computing Research Institute, I was involved in projects concerning the domain of Urban Computing and Data Science.

1. Developed a demo-application to estimate veracity of online news data. The application was presented at the World Wide Web Conference '17.
2. Developed a tool to estimate pollution in the city of Doha using traffic density data.
3. Worked on a project to correlate sensor data with social media data to explore the concept of connected cities.

Collaborators: Dr. Laure Berti-Equille, Dr. Mouhamadou Lamine Ba, Dr. Sanjay Chawla

• Date: Jan 2016 - August 2016

Close

Data 2 Information

---

Working as a Decision Scientist at InnovAccer, I worked with in a team of 4 to develop an application to scan digitized documents such as Ration Cards. I wrote back-end infrastructure of the program using Python, Flask, Tesseract-OCR and OpenCV

• Date: September 2015 - December 2015

Close

Mining Society Beliefs about Gender-based Violence

---

I did my undergraduate thesis remotely with KNO.E.SIS, where I developed a tweet classification to mine opinon of people around gender-based violence on social media.

I had developed a system in Java and Weka by following the classicial text analysis steps - Data Labelling, Tweet Labelling, Undertaking Inter rater reliability check, Data Cleaning, Feature Extraction, Training Model, Prediction!

• Date: August 2015 - May 2015

Close