Skip to content
View Tahsin-Mayeesha's full-sized avatar

Block or report Tahsin-Mayeesha

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Please don't include any personal information such as legal names or email addresses. Maximum 100 characters, markdown supported. This note will be visible to only you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
Tahsin-Mayeesha/README.md

Hi 👋, I'm Tahsin-Mayeesha. I'm working as a researcher focused on NLP, HCI and AI policymaking 🌍 ⚖️ 🕊️ 🧑‍⚖️. Currently, I am diving into exciting new challenges at EBLICT-Dream71 joint collaboration on Bangla VPA (Virtual Private Assistant) project funded by the Bangladesh government where we build Llama based RAG models for making public administration related information accessible to citizens.

Prior to this role, I worked as a HCI researcher for a year in Design,Inclusion & Access Lab(DIAL) in North South University with my superviser Dr. Nova Ahmed in multiple projects in education and AI policy-making. The key projects are 1)"My Freedom in Light", where we investigated challenges of women in computing with fictional inquiry and co-design workshop faciliating participatory design from female students in computer science, and 2) "Designing Accountable and Ethical AI for the Next Billions: Considering the Needs of South Asian Marginalized Communities" - where we studied current state of AI ethics related policies and perspectives with a focus on South Asia and Bangladesh . Publications from these projects have been accepted to ICTD 2024, ACM Compass 2024 and Ubicomp 2024.

Simultaneously, I worked on NLP research as a senior research assistant with supervision of Dr. M. Rashedur Rahman in the project "Bengali NLP : Application in Literature and Natural Language Generation" on multiple topics including answer-aware question generation from passages, and generating questions with/without guidance from images. I also finished a predoctoral fellowship in Fatima Fellowship on a NLP project with mentor Benjamin Muller on investigating cultural biases like formality in multilingual generative models. Publications from these projects were accepted to journals and NLP Conferences (EMNLP 2023, MM-NLG Workshop held with INLG 2023).

I've graduated from Computer Science and Engineering (North South University) in 2020. My thesis project was on building deep learning models for question answering systems in Bengali where I trained multilingual BERT models on synthetic data. During my undergrad I worked with Tensorflow Hub team for Google Summer of Code 2019 with mentor Vojtech Bardiovský, Berkman Klein Center of Internet and Society with mentor Hal Roberts for Google Summer of Code 2018 and Cramstack in 2017.

When I am not immersed in research, I enjoy watching anime, reading manga or books, and taking care of my cats.🐈‍⬛

tahsin-mayeesha

Connect with me:

tahsin_mayeesha tahsin-mayeesha mayeesha mayeesha.tahsin tahsin_mayeesha @tahsin_mayeesha

tahsin-mayeesha

 tahsin-mayeesha

Pinned Loading

  1. BanglaQG BanglaQG Public

    Source code for Bengali Question Generation Paper

    Jupyter Notebook 1

  2. Mediaviz Mediaviz Public

    Force Based Network Visualization Library With Automated Scaling To Prevent Node Overlap, Label Adjustment & Easy Community Visualization Created During Google Summer Of Code 2018 With Berkman Klei…

    Python 15 8

  3. gsoc2019 gsoc2019 Public

    Work with Tensorflow Hub on Google Summer of Code 2019. Features prototype ULMFiT implementation, pretrained embedding exporter and a Bangla text classification notebook.

    Jupyter Notebook 4 5

  4. Dhaka_Tribune-Network-Visualization Dhaka_Tribune-Network-Visualization Public

    Network visualization of media coverage of violence against women in Bangladesh, featured by fast.ai, write-up in the provided link

    JavaScript 4

  5. anime-recommender anime-recommender Public

    Content based simple anime recommender from Myanimelist Dataset on Kaggle

    Jupyter Notebook 3

  6. bangla-fake-news bangla-fake-news Public

    mbert trained on banfakenews dataset. winning blog post in weights and bias study group competition. model served in huggingface hub.

    Jupyter Notebook 1