This project Inclusive.AI introduces a paradigm for diverse stakeholders of AIto collaboratively address social, technical, and political challenges. This is a step foward to democratize AI governance by ensuring participation in decision-making regarding AI models and behaviors through a Decentralized Autonomous Organization (DAO) mechanisms. This includes various decision-making schemes like quadratic, ranking, and weighted voting, with both equal and differential voting powers. Leveraging blockchain technology, Inclusive.AI promotes transparency and prevents manipulation, enhancing user involvement in governing large language models and shaping future policies.
Tanusree Sharma Jongwon Park, Yujin Kwon,
Yun Huang, Sunny Liu, Dawn Song, Jeff Hancock, Yang Wang. Inclusive.AI: Engaging
Marginalized Groups in Democratic Decision-Making in AI. [Under Review]
[Email me for Preprint]
Tanusree Sharma, Yujin Kwon,
Kornrapat Pongmala, Henry Wang, Andrew Miller, Dawn
Song, and Yang Wang. "Unpacking How Decentralized
Autonomous Organizations (DAOs) Work in Practice
[Under Review CHI 2024] Preprint version
Tanusree Sharma*, Silvan Bruggmann*,
Robin Fritsch, Jianing Si, Roger Wattenhofer, Ye Wang.
From Discussions to Decisions: Understanding User
Behavior in Decentralized Autonomous Organizations. [Under Review CHI 2024]
Tanusree Sharma*, Smit Desai*, Pratyasha Saha*. Using ChatGPT in HCI Research—A Trioethnography
ACM CUI 2023 PDF
This line of work proposed new method of dataset creation and designing AI for underserved group (e.g. blind users)in ethical way. A novel contribution of this work is an addition of new data creation/collection method for ethical AI development pipeline - opting for props objects to create dataset with blind users; integrating diverse expertise areas like accessibility, privacy, and computer vision to ensure essential criteria like content diversity and category representation as well as privacy/security of data collection/creation to foster AI advancements. Dataset comprises 728 images and 728 videos from 14 private categories, all captured by 26 visually impaired individuals. Furthermore, this work contribute an initial prototype and evluation of blind users with design insights of how blind users would like to obfuscate private content in different contextual scenarios, complemented by a Wizard-of-Oz setup offering human-generated guidance.
The BIV-Priv Image dataset is available
here
The BIV-Priv Video dataset is available
here
Tanusree Sharma Abigale Stangl,
Lotus Zhang, Yu-Yun Tseng, Inan Xu, Leah Findlater,
Danna Gurari, and Yang Wang. "Disability-First Design
and Creation of A Dataset Showing Private Visual
Information Collected With People Who Are Blind." In
Proceedings of the 2023 CHI Conference on Human
Factors in Computing Systems
[PDF].
Yu-Yun Tseng, Alexander Bell,
Tanusree Sharma, Lotus Zhang, Abigale
Stangl, Leah Findlater, Yang Wang and Danna
Gurari. Locating Objects in Images Taken by People With
Visual Impairments. [Under Review International Journal of Computer Vision (VISI)]
Lotus Zhang, Abigale Stangl,
Tanusree Sharma, Yu-Yun Tseng, Inan
Xu, Leah Findlater, Danna Gurari, and Yang
Wang. Designing Accessible Obfuscation Support for Blind
Individuals’ Visual Privacy Management. [Under Review
CHI].
Smirity Kaushik, Natã M. Barbosa, Yaman Yu,
Tanusree Sharma, Zachary Kilhoffer, JooYoung Seo,
Sauvik Das, and Yang Wang.. GuardLens: Supporting Safer Online
Browsing for People with Visual Impairments. Symposium on Usable Privacy and Security (SOUPS 2023)
PDF
This line of work includes usable privacy and security addressing investigating the context of Blockchain and decentralized application for addressing issues around both end users and developers' points of view. One project presented at Usenix 23 dedicated to improving smart contract development practices and designing tools for secure coding of smart contracts. Meanwhile, another project focuses on exploring the security and usability aspects of web3-based applications, presented at SOUPS 23, including crypto wallets and DeFi applications. Another noteworthy project involves addressing usability and security concerns by implementing cryptographic protocols like MFKDF in the context of crypto wallets with a potential to benefit various decentralized applications These project applied ranges of method including, system developemt, randomized control experiment, user interview, surveys to understand users expectation, preferences in making actionable tooling/application. As a result of the efforts, we have developed two exceptional systems: the Ciao wallet with the MFKDF scheme and the i-Wallet with enhanced usability and accessibility features.
Tanusree Sharma,, Zhixuan Zhou, Andrew Miller, and Yang Wang. A {Mixed-Methods} Study of Security
Practices of Smart Contract Developers. In 32nd USENIX
Security Symposium USENIX Security 23
[PDF]
Tanusree Sharma*, Zhixuan Zhou*, Luke Emano, Sauvik Das, Yang Wang., Iterative Design of An Accessible
Crypto Wallet for Blind Users. Symposium on Usable Privacy and Security SOUPS 2023
PDF
Tanusree Sharma, Chixuan Zhou, Yun Huang, &
Yang Wang.It's A Blessing and A Curse": Unpacking
Creators' Practices with Non-Fungible Tokens (NFTs)
and Their Communities. Crypto Economics Security
Conference (CESC)
PDF
Tanusree Sharma*, Vivek Nair*, Henry
Wang, Yang Wang, and Dawn Song “I Can’t Believe It’s
Not Custodial!”: Usable Trustless Decentralized Key
Management. [Under Review CHI 2024].
Tanusree Sharma*, A.T.M Mizanur
Rahman*, Silvia Sandhi, Yang Wang, and S M Taiabul
Haque. Confused: Navigating the Current Practices and
Perceived Risks among the Cryptocurrency Users in
Bangladesh. [Under Review CHI 2024].
Yaman Yu, Tanusree Sharma, Sauvik
Das, and Yang Wang "Don't put all your eggs in one
basket'': How Cryptocurrency Users Choose and Secure
Their Wallets. [Under Review CHI 2024].
Zhuangtong Huang, Rui Jiang,
Tanusree Sharma, Ye Wang. Changing
Trust: Auditing's Impact on Web3 Security from a User
Perspective. [Under Review CHI 2024].
Yongqi Guan, Yaman Yu,
Tanusree Sharma, Zhuangtong Huang,
Rui Jiang, Kaihua Qin, Yang Wang, and Ye Wang. Across
Fiat and Cryptocurrency: User Security Perceptions in
Stablecoin Practices.[Under Review CHI 2024].
Jianing Si, Tanusree Sharma, Yang
Wang, and Ye Wang. Understanding User-Perceived
Security Risks and Mitigation Strategies in the Web3
Ecosystem. [Under Review CHI 2024].
This line of work is motivated by the cultural, economic, and development differences in the Global South, where the Western definition of privacy is often not applicable. As a result of this project, we found novel user mental models of targeted ads based on mobile app permissions and excessive AI usage. People from Global South often preferred ad control over transparency. Another series of project under this line of work involves exploring and designing ad technology, emerging crypto technology, AI, etc, and technologies used in crisis situation, such as COVID-19.
Tanusree Sharma ,Smirity Kaushik, Yaman Yu, Syed Ishtiaque Ahmed, Yang Wang.
User Perceptions and
Experiences of Targeted Ads on Social Media Platforms:
Learning from Bangladesh and India. CHI Conference on Human Factors in Computing Systems (CHI23)
PDF
Tanusree Sharma,, Ankit Shrestha *, Pratyasha
Saha, Syed Ishtiaque Ahmed, Mahdi Nasrullah Al-Ameen.
A First Look into Software Security Practices in Bangladesh. ACM Journal on Computing and Sustainable Societies. ACM COMPASS 2023
PDF
Tanusree Sharma, Md Mirajul Islam, Anupam Das, SM Taiabul Haque, and Syed Ishtiaque Ahmed.. Privacy
during Pandemic: A Global View of Privacy Practices
around COVID-19 Apps. In ACM SIGCAS Conference on
Computing and Sustainable Societies. ACM
Compass 2021
PDF
Tanusree Sharma* , A.T.M Mizanur
Rahman*, Silvia Sandhi, Yang Wang, and S M Taiabul
Haque. Confused: Navigating the Current Practices and
Perceived Risks among the Cryptocurrency Users in
Bangladesh. [Under Review CHI 2024].
Data is a valuable asset for organizations in data-driven economies, but without effective measures, data misuse can negatively impact end-users. The 2018 GDPR gives individuals control over their personal data. This project aims to explore users' understanding of data minimization in search engines, factors influencing their decision to share search data, and their perspective on how data minimization should work in AI System.
Tanusree Sharma, Lin Kyi, Yang Wang,
Asia Biega (To appear Usenix 24).
Tanusree Sharma. Rethinking Data
Minimization from a {User-Centered} Approach: A
Paradigm Shift. Usenix PEPR 2023
PDF
Tanusree Sharma, Wang, T., Giulio, C. D., & Bashir, M.
(2020, October). Towards Inclusive Privacy Protections
in the Cloud. In International Conference on Applied
Cryptography and Network Security (pp. 337-359).
Springer, Cham
PDF