Projects
I 💜 building cool stuff 👨💻 💻
Here are some of my projects & publications
-
Revisiting Open Domain Query Facet Extraction and Generation| ICTIR’22
This paper discusses various approaches to open domain query facet extraction and generation including sequence labeling and extreme multi-label classification. Proposed methods outperform state-of-the-art baselines in both precision and recall, and a toolkit named Faspect is released to include various model implementations for this task.
-
Maruna Bot: An extensible retrieval-focused framework for task-oriented dialogues
This report presents the first iteration of Maruna Bot, a Task-Oriented Dialogue System (TODS) that assists people in cooking or Do-It-Yourself (DIY) tasks using either a speech-only or multimodal (speech and screen) interface. This work was done as part of the Alexa Prize Taskbot challenge Proceedings 2021.
-
This paper introduces the novel dataset Covid-HeRA (Covid Health Risk Assessment). This work studies the severity of each misinformation-related social media post, i.e., how harmful a message believed by the audience can be and what type of signals can be used to discover high malicious fake news and detect refuted claims.
-
Senior Thesis: Deep Patient for Summarizing Health Records
Recreated the Deep-Patient architecture that generates an embedding for EHRs by using stacked deonising autoencoders. This out-performs other predictions based on raw-data and other dimensionality-reduction methods.
-
Image Similarity using Deep Ranking
Re-implementation of paper, “Learning Fine-grained Image Similarity with Deep Ranking”. Trained ResNet101 neural net on TinyImageNet across multiple GPUS.
-
A Web app that connects video segments across various online courses. The segments are connected by the similarity of concepts covered in them. I implemented and experimented with various text analysis algorithms such as Semantic similarity and TF-IDF scoring to make a similarity metric.