top of page

The theme of my work is "designing for recognition." Historically recognition has remained one of the empowerment agendas of marginalized communities. I build on the social science theories of recognition and argue for four design orientations for computing technologies for recognition: overcoming colonial oppression, sustaining justice, addressing local moral value systems, and combating opaque obstacles. Therefore, in my research, I worked on decolonizing data, algorithm, and visualization practices in the Global South, designed for transformative justice, had a critical look at and build connections among faith, myth, and misinformation, and designed to address fear, stigma, and exclusion. I discuss my projects under the following four themes:

Decoloniality-data-algorithm-visualization.png

A witch in Kazipur village showed several example Jantras used for Tantra in their witchcraft practice. (Clockwise from left) Jantra for solving marriage issues, fo witctcraft training, for dismissing effects of bad spells, and for better health.

katha_22_edited.jpg

A rural women from Mandartola village showed us her great-grandmother’s Nakshi-Katha explained that the wheel from the top middle of the main Nakshi-Katha what has the monetary records of 16 of the productive lunar cycles (8 months) that year,

Decolonizing Data, Algorithm, and Visuals in the Global South

paper clip_edited_edited_edited.png

The objective of this project is to decolonize data, algorithm, and visualization in the Global South using local intelligence, which includes faith-based beliefs, myth, language, and arts and crafts. I study the knowledge, materials, and politics involved in rural Bangladeshi local art, culture, visuals and computing in rural witchcraft, Nakshi-Katha, and Hindu idol-making practices and find how HCI research may benefit by making a deeper engagement with various local moral values, emphasizing communal relationships, and neutralizing radicalism.

Selected publications and activities: 

  • [CHI 2023Abstraction and semiology in constructing visual narratives,

  • [FAccT 2022] Big data and AI in Global South [Workshop], 

  • [Mozilla Festival 2022] Decolonizing AI [Workshop],

  • [ECSCW 2022] Fieldwork reflections on pictorial consent,

  • [CHI 2021] Understanding luck, hunch, faith, and data-driven prediction in betting

  • [CSCW 2021] Rural fact-checking procedure for misinformation [🏅Diversity and inclusion recognition]

  • [CSCW 2020] Contrasting grammar of modern data visualization and local traditional visualization

  • [ACM Interactions 2020] Rural faith-based practice in combating COVID-19,

  • [CHI 2019] Witchcraft and HCI: Morality, modernity, and postcolonial computing.

Work-flow diagram of Unmochon application with three major components: plugin at the user-end, the server with end-to-end encryption, and the customized Facebook group.

I examine the opportunities and issues that arise in designing technologies to support low-income rural women in Bangladesh. I conduct qualitative and quantitative studies to explore the systemic everyday challenges women face that form the backdrop against which technology design could potentially happen. In this ongoing research, I investigate women's harassment on Social Media and real life and design technologies to support gender justice.

This project is supported by Facebook/Meta Fellowship and ICT Innovation Grant by ICT Ministry, Bangladesh, 2020

Selected publications and activities: 

  • [CHI 2022] ShishuShurokkha: A transformative justice approach to combat child sexual abuse,

  • [CHI 2021] Unmochon: A tool to combat online sexual harassment by shaming [🏅Best paper honorable mention], 

  • [COMPASS 2021] Gender-selective access to information for women, 

  • [ICTD 2019] Shada-Bakso: A tool for women empowerment for fighting the fear of technology,

  • [CHI 2018] Opportunities and challenges of designing within a patriarchal society [🏆 Best paper award]. 

Media and Press:

Design for Transformative Justice

Access points for design action in women’s lives. The most significant relationships that shape women’s lives are the relationships with husbands and in-laws indicated in dark gray. The width of an arrow indicates the frequency with which interactions happen. The length of an arrow gives a sense of the intimacy of the relationship. Dotted arrows indicate relationships rarely drawn on. The gray arc cuts through relationships that are most significantly mediated through husbands and in-laws.
Unmochon
paper clip_edited_edited_edited.png

Access points for design action in women’s lives. The most significant relationships that shape women’s lives are the relationships with husbands and in-laws indicated in dark gray. The width, and length of an arrow indicate the frequency of interactions and intimacy of the relationship.

A typical rural tea-stall in Baulia village where village men have gathered with an intention to bet on cricket match. They follow the cricket match updates on their expert's mobile phone and set their bets. 

A bird-shaped Jantra in the middle having Mantra/Dua written all over it. It is a hybrid one having both Islamic scripts and Kali-Mantra. A Knife, wax, and black thread are being used it in a Tantra. Photo taken in Bejpara village. 

Faith, Myth, and Misinformation

65006812_917861338558050_7595724133413945344_n.jpeg
w3.jpg

I study the values embedded in mis(information), rural computing, and wellbeing practices. Through my work with rural Bangladeshi people, I inform HCI-design and social wellbeing research about how to design more usable and appropriate technologies by integrating cultural values and local intelligence. 

Selected publications and activities:

  • [CHI 2022] Integrating faith, religion, and spirituality in HCI [workshop],

  • [CSCW 2021] Rural fact-checking procedure for misinformation [🏅Diversity and inclusion recognition]

  • [COMPASS 2021Oshudh Poro: A mobile phone application for personal medication management,

  • [alt.CHI 2020] Parareligion, alternative rationality, and HCI,

  • [CHI 2020] Islamic-HCI: Designing with and within Muslim populations [workshop],

  • [ACM Interactions 2020] Combating COVID-19 with faith,

  • [CHI 2019] Witchcraft and HCI: Morality, modernity, and postcolonial computing,

  • [CSCW 2019] Faith-based and situated knowledge in rural healthcare practices.

Media and press:

  • Eurasia Review: Rural Bangladeshis turn to faith, family for fact-checking

  • Cornell Chronicle: Rural Bangladeshis turn to faith, family for fact-checking

  • Phys.org: Rural Bangladeshis turn to faith, family for fact-checking

  • Opera News: Rural Bangladeshis turn to faith, family for fact-checking

  • News Wise: Rural Bangladeshis turn to faith, family for fact-checking

paper clip_edited_edited_edited.png

(Left) a participant from Shankorpur village is using Shada-baksho, a system we designed to help rural people get over their fear of technology and make calls on smartphones. (Middle and right) application interfaces of Shada-baksho: pages showing call logs and contact saving options.

(Left) a bettor in Kochua village is following cricket match updates and betting rates online and (right) doing his calculations before he bets.

I explore the social and cultural factors that influence the online betting practices among the villagers and how bets harmonize with users’ faith, hunch, and cultural practices, along with statistical recommendations. Shada Bakso is a hardware device, designed by us, to explore the fears of using mobile phones among the rural women of Bangladesh. My probe study with the rural women using Shada Bakso suggests that rural women's fear of technology further initiates technology non-use.

Selected publications and activities: 

  • [CSCW 2022] Imagined communionship, sovereignty, and inclusiveness in Facebook Groups,

  • [CSCW 2022] Kabootar: Informal and trustworthy Fintech for economically sanctioned communities,

  • [ICTD 2022] Decorum, civility, and certainty for toleration on social media,

  • [COMPASS 2022] Intermediaries and terrorist assemblage in cashless transactions,

  • [COMPASS 2021] Gender-selective access to information for women,

  • [COMPASS 2021] Oshudh Poro: A mobile phone application for personal medication management,

  • [COMPASS 2020] A survey on social media privacy concerns in the Global South, 

  • [ICTD 2019] Shada-Bakso: A tool for women empowerment for fighting the fear of technology

  • [Third-Space Lab, University of Toronto] Orality and mobile money [workshop].

Fear, Stigma, and Exclusion

Screen Shot 2021-09-16 at 05.14.08.png
p4.png
paper clip_edited_edited_edited.png
paper clip_edited_edited_edited.png
Screen Shot 2021-11-26 at 14.00.34.png
Screen Shot 2021-09-15 at 11.44.18.png

(Top) flow diagram of the process of speech signal processing and feature extraction. (Bottom) hierarchial tree diagram of speech emotion recognition.

Bar charts comparing the accuracies of speech emotion recognition algorithms developed in this series of iterations. Our research achieved a maximum of 80.55% accuracy rates in emotion recognition from Speech data.

Previously I worked on emotion recognition using speech dataset (SER). In this work, I investigated many popular forms of speech features including wavelet coefficients, entropy, and mel-frequency cepstral coefficients (MFCC). I used both hierarchical and non-hierarchical approaches to apply prominent classification algorithms such as K-nearest neighbors (KNN), K-means clustering, Euclidian distance (ED), and support vector machine (SVM). I concluded that designing a speaker-independent SER system should focus on speaker-independent features of speech signals. My frequency-based analysis provided significantly higher accuracy than many of the state-of-art works. My algorithm produced 80.55% correct results while the best-known previous algorithm had 56.98% of success.

 

Selected publications: 

  • [DSP 2015] SER from EMD wavelet using KNN, 

  • [ISCAS 2014] SER from teager energy (TE) operated entropy of wavelet coefficients using ED

  • [MWSCAS 2014] Enhanced TE-wavelet coefficients for SER using K-means and KNN

  • [ICIEV 2014] A non-hierarchical SER using K-means,  

  • [ICEEICT 2014A hierarchical SER using ED .

Speech Emotion Recognition (SER)

bottom of page