👩‍🏫 Undergraduate Teaching Assistant

• Assisting Professor in Problem Solving in Python with over 70 students to enhance student understanding by holding regular office hours and providing mentorship.
• Help professor grade and give feedbacks to students’ homework and programming assignments.
• Holding weekly recitations to answer students’ questions about lectures materials.

🥼 Undergraduate Research Assistant

• Assist Professor Pustejovsky’s Lab to complete domain specific AMR (Abstract Meaning Representation) annotation.
• Manually annotate and validate machine generated AMR graphs using AMR-dict and UMR writer.
• Design and propose domain specific rules to adopt AMR graphs to suit for recipe data set
• Using Python, spaCy and Flask to process jsonlines file and visualize event trigger and entity spans across sentences.
• Generating GLAMR that incorporates subevent structure for predicates and designating opposition structure for the object undergoing change.

🌐 Member of Branda

• Collaborated with other team members to develop and optimize existing features of Branda app for Brandeis University
• Designed new layout of Special Event section for Branda APP that can display Brandeis upcoming events and direct to the relevant website
• Focusing on Frontend and improved features in the mobile app using Figma and React native

🏃‍♀️ President of Brandeis Running Club

• Created Brandeis first Running Club and recruited more than 100 members
• Organize running events twice a week, manege club’s budget and plan for semester events
• Managed meetings with e-board members to set goals and stimulate enthusiasm of runners in Brandeis

🤔 SWiT: Emoji Annotation

Emojis have surged in popularity every since their invention. People use them to add more emotions to the plain text.
However, emojis also have their unique connotation: the same emoji can have completely different meanings depending on the context in which it appears.
Our emoji annotation project aims to understand the nuanced connotations of three popular emojis:

1.😭 (loudly crying face)
2.🥲 (smiling face with tear)
3.🥺 (face holding back tears)

We designed the annotation guidelines to annotate the emoji emotions and we ran sklearn’s Logistic Regression model and sklearn’s Naive Bayes model on it to get a baseline.

📲 Emoji Generator for Texts

Emoji project again 🥳
There are tons of emojis, but when we are messaging, it takes forever to find the perfect emoji suited for our text 😫.
Our goal is to augment sentences with suitable emojis to imbue it with liveliness and emotion. And we designed a webpage using Streamlit to deploy this functionality

For example:
Input: “I like noodles”
Output: ❤️🍜 (suggested emoji by the model)

👩‍💻 HackMIT project

Collaborated with team members and investigated Audio Digital Signal Professing and Natural Language Processing techniques.
I also designed a website integrated transcription, summary, and question generation using Figma.
And this website can provide functions of audio to text translations, summary generation, and QA generation using transformer, HEPOS, entity highlighting.