Hi, I'm Aidan!

  • I'm a fourth year PhD student born in Virginia and studying at the University of Virginia.
  • I'm happily studying Natural Language Processing advised by Prof. Yangfeng Ji.
  • I'm grateful to be fully funded by the Dean's Scholar Fellowship (3 years) and the Distinguished Fellowship.
  • I received my undergraduate degree from the University of Illinois at Urbana-Champaign Magna Cum Laude in Computer Science + Linguistics, and my MCS at UVA.
  • I'm currently working on the summarization of factchecks, and I have a strong interest in the interpretability of NLP models and applications of NLP for social good.
  • In my freetime, I enjoy playing guitar, playing boardgames, and cooking.

Contact Info:

Email: aidan.w.sanPLEASEIGNORETHISPART@gmail.com


Twitter: @aidanwsan
Pronouns: he or they (no preference)

Research:

MOOC Concept Detection (Learning @ Scale - Demonstration Session 2020)
  • Detecting concepts (ie. Addition, Arithmetic, Calculus) in MOOC courses
  • Concept Chunker - Uses IOB tagging to determine where multiword concept phrases start and end
  • LSTM neural network trained on textbook data
  • Results on MOOC Lecture Dataset (Test): Accuracy: .933, F1: .62
  • ACM DL, GitHub


PLAtE - Web Extraction dataset
  • Dataset of more than 50k items for automatic extraction of product information from websites
  • Wrote handcrafted rules to extract product attributes, then performed cleaning using crowdsourced annotations
  • Evaluated dataset using a SOTA neural web extraction model
  • arXiv


Irony Detection (SemEval 2018)
  • 2 Stacked LSTMs which incorporate Emoji Embeddings and Sentiment Scores
  • Rank 6/31 (F1 score) for subtask B (multitask)
  • ACL Anthology, GitHub

Positions:

  • CS1110 Instructor (Spring22) - Primary lecturer for one section of Intro to Programming
  • Amazon Applied Science Intern (Summer21) - Designed a dataset to train neural web extraction models on the Alexa AI team
  • Grail SWE Intern (Summer19) - Wrote Go code to allow switching of instruments during assays on the automation team
  • Facebook SWE Intern (Summer18) - Wrote Python code to improve translation data quality on the Applied Machine Learning team
  • TA for Discrete Math (Fall20), Natural Language Processing (Spring21), Data Structures (Fall17-Fall18)
  • Space Chair of UVA CSGSG (Spring19-Fall21) - Managed budget for the CS lounge and organized events
  • Cofounder of the UVA NLP Reading Group (Fall19-Spring20)
  • Founder/Chair of ACM Special Interest Group for Natural Language Learning (Fall18-Spring19) - Developed course material about NLP topics