Short Reports

Monitoring War Destruction from Space Using Machine Learning

Arogya Koirala

Extracting information on-war related destruction is difficult because it relies on eyewitness accounts and manual detection on the field, which is not only costly for institutions carrying out these efforts, but also unsafe for the individuals carrying out this task. The information gathered is also incomplete, which makes it difficult for use in media reporting, carrying out humanitarian relief, understanding human rights violations, or academic reporting. This seminar introduces an automated approach to measure destruction in war damaged buildings by applying deep learning in publicly available satellite imagery. We adapt different neural network architectures and make them applicable for the building damage detection use case. As a proof of concept, we apply this method to the Syrian Civil War to reconstruct war destruction in these regions over time. We close the discussion by talking about how the nature and quality of the inputs used (publicly available satellite imagery) and different architectural choices made in the design of the machine learning system relate to the robustness and generalizability of the outputs produced. This work builds on prior work by Mueller et al. in the PNAS paper "Monitoring war destruction from space using machine learning".

Modernizing Mainframe RACF

Shai Dhaliwal

I plan to explore how cloud modernization will improve cyber security for legacy mainframe information systems: Consider the driving factors which influence organizations to leverage legacy mainframe systems today and existing capabilities for Identity & Access Management; explore opportunities for organizations to migrate IBM Mainframe workloads securely to the cloud and how this could benefit organizations for the next 10-20 years, and provide a recommended framework to execute Mainframe RACF modernization for improved Identity & Access Management security.

Exploring Consumer Robocall Mitigations

Calvin Lee

Have you recently received a strange call from an unknown number in your area code requesting that you extend your expired vehicle warranty insurance? If you have, you had one of the 75.9 billion illegal robocalls that were made within the last 12 months. Counter-advances have been made by critical players such as the Federal Communications Commission and Tier-1 US network carriers, but are they proving futile? Perhaps we can look towards similar tactics such as email anti-spam or CAPTCHA techniques. Designing a solution to mitigate these unwanted calls is complex but we will explore new possible mitigations.

Drawing Lines Between Section 230 and Trust & Safety

Alan Kyle

In this presentation I will discuss what Section 230 and trust & safety are and how they are connected. I will use my experience as a trust & safety professional, and my research of Section 230 bills as a jumping off point for thinking about ways to contextualize current attempts to regulate the Internet.

English Documentation of Non-English Stories: A Study of the People’s Archive of Rural India

Siddharth Adelkar

The People’s Archive of Rural India (PARI) is primarily a journalism project. PARI reports on the "everyday lives of everyday people" by using the craft and structure of journalism to inform their predominantly English speaking urban readership — an important part of which is school and college students. PARI sees itself as an antidote to the structural problems in Indian media, education and historiography. While Indian media focuses on the urban rich, PARI focuses on the rural poor. While Indian English education focuses on skilling and emigration, PARI focuses on deep engagement with one’s surroundings. While Indian history focuses on the narrative of kings and empires, PARI prepares an archive for future writings of people’s histories. The unique themes that have stood out in PARI’s reportage include the agrarian crisis in India, women’s sexual and reproductive health, and chronicling the impending climactic catastrophe as "everyday" people feel it.

An important question that I wish to study in PARI’s coverage is the impact — benefits and drawbacks — of English-first documentation of livelihood and culture in a predominantly non-English speaking country. PARI is translated in up to 11 Indian languages by over 100 accomplished translators. Yet, in the future if only PARI were to survive what will be lost due to the English-ness or English-firstness of the stories? Will something be gained? Does the presence of pictures and video improve the situation or worsen it?

The ‘Fungibility’ of Non-Fungible Tokens: Vulnerabilities in an Over-Hyped Market

Sarah Barrington

Non-fungible tokens, digital certificates of ownership for virtual art, are becoming increasingly ubiquitous in Western media; from releasing the first notes of the Beatles’ ‘Hey Jude’ as an NFT, to the first tweet being sold as an NFT for $2.9 million. In 2021, the market was valued at $17.6bn, representing a growing and salient portion of the overall cryptocurrency and blockchain economy. Yet, the NFT market is also speculative, variously described as irrational and overhyped. The emergence of vulnerabilities, along with a sustained market downtime, are now calling the role of NFTs into question: what exactly are NFTs? And most importantly, what gives them value? This project aims to address these questions, arguing that three fundamental properties (permanence, immutability and uniqueness) are necessary (but not sufficient) conditions for an NFT to have value. We explore both the underlying artworks and their associated metadata in order to define these metrics. Furthermore, we take a quantitative approach to testing these definitions against 6 months of real-world data, examining the true permanence and perceived value of over 7 million NFTs. We ultimately envision this work to help buyers and marketplaces identify and warn users against purchasing NFTs that may be overvalued, and bring some much needed rigour to a presently complex and recondite market.

Assessing Data Subject Access Requests

Ameya Naik

Any mobile or desktop application prompts you to sign a term of policy agreement that allows the application to gather information related to you, your activities, and your attributes. The level and type of information gathered depend on the organization, business model, kind of application, and geographical location. You could access this information through Data Subject Access Requests, which the applications are bound to provide. While GDPR (Article 15) and CCPA have broadly drawn rules and regulations for Data Subject Access Request(DSAR), however, there still are differences in the way the data is stored and shared with the consumers(you). I plan to initiate and note the process of DSARs and then analyze the data shared. This analysis could potentially lead to interesting observations, and would want to compare the shared data by similar applications (messaging, social media etc.). Developing a catalog and visualizing the data would be the other aspect of the project.

This seminar will be held both online & in person. You are welcome to join us either in South Hall or via Zoom.

For online participants

Online participants must have a Zoom account and be logged in. Sign up for your free account here. If this is your first time using Zoom, please allow a few extra minutes to download and install the browser plugin or mobile app.

Join the seminar online