Eleanor M. Lin
elealin@umich.edu
GitHub
Google Scholar
LinkedIn
I am a first-year Computer Science and Engineering PhD student at the University of Michigan, advised by David Jurgens. I obtained my BA at Columbia University, double-majoring in linguistics and computer science. I am primarily interested in natural language processing. At Columbia’s Speech Lab, I researched code-switching under the mentorship of Professor Julia Hirschberg. I have also researched the use of large language models for the study of negotiation at the University of Southern California’s Affective Computing Lab, where I was advised by Professor Jonathan Gratch. Previously, working with Professor Vicente Ordóñez Román in the Vision, Language, and Learning Lab at Rice University, I investigated the relationship between images’ visual complexity and image descriptions’ linguistic complexity. Working with Dr. Kate Moore in the Corter Lab at Columbia’s Teachers College, I have also researched communication patterns in collaborative learning. You can find my curriculum vitae here.
- Publications
- Presentations
- Grants, Awards, and Academic Honors
- Research Experience
- Projects
- Teaching Experience
- Work Experience
- Science Communication
- Service
Publications
Refereed Publications
-
RASwDA: Re-Aligned Switchboard Dialog Act Corpus for Dialog Act Prediction in Conversation
International Workshop on Spoken Dialogue Systems Technology. Sapporo, Japan. Accepted as poster presentation. [Paper]
-
Toward a Better Understanding of the Emotional Dynamics of Negotiation with Large Language Models
In The Twenty-fourth International Symposium on Theory, Algorithmic Foundations, and Protocol Design for Mobile Networks and Mobile Computing (MobiHoc ’23), October 23–26, 2023, Washington, DC, USA. ACM, New York, NY, USA, 6 pages. Accepted as oral presentation. https://doi.org/10.1145/3565287.3617637 [Paper] [Slides] [Poster]
Thesis
-
Developing Performant Models for Translating Spoken Taiwanese Into Spoken English Using Free and Publicly Available Resources
Columbia University Program of Linguistics, April 2024. Undergraduate thesis. [Thesis]
Working Papers and Preprints
-
Switching Tongues, Sharing Hearts: Identifying the Relationship between Empathy and Code-switching in Speech
Accepted to INTERSPEECH 2024; *DB and EL are equal contributors to this work and are designated as co-first authors.
-
Detecting Empathy in Speech
Accepted to INTERSPEECH 2024.
Presentations
-
Instilling discipline in LLM-based negotiating agents
Annual Conference of the International Association for Conflict Management (IACM), Singapore, June 2024. Symposium talk.
-
Toward a Sociolinguistic Profile of Taiwanese Americans
Emory Undergraduate Linguistics Conference, Online, April 2024. Oral presentation. [Working Paper] [Slides]
-
Text-Based Prediction of Visual Complexity: How Does What We See Influence What We Say?
Columbia University Undergraduate Research Symposium, New York, NY, October 2022. Poster presentation. [Report] [Slides] [Poster] [Code] [Project Blog]
Grants, Awards, and Academic Honors
-
Departmental Honors in Linguistics, Columbia University Program of Linguistics, 2024
One of two graduating linguistics majors selected for honors, on the basis of the quality of the senior thesis.
-
Computer Science Scholarship Award, Columbia University Department of Computer Science, 2024
One of two B.A. candidates selected annually on the basis of “outstanding academic achievement in computer science."
-
Fulbright-Hays Group Projects Abroad (GPA) Fellowship for Advanced Overseas Chinese Language Study, American Councils for International Education, 2024
Awarded $7000 to cover cost of attendance for Taiwan Intensive Summer Language Program.
-
Dean's List, Columbia College, Expected 2020 - 2024
-
Work Exemption Program Grant, Columbia University, 2024
Recipient of $1,820 Columbia Grant to support pursuit of research in the Columbia Speech Lab.
-
Outstanding Undergraduate Researcher Award - Honorable Mention, Computing Research Association, 2023
-
Junior Phi Beta Kappa, Columbia University, 2023
Awarded to 2% of the Columbia College Class of 2024 (26 students) in recognition of breadth, depth, and rigor in scholarship
-
Upsilon Pi Epsilon (International Honor Society for the Computing and Information Disciplines) Member, Columbia University, 2023
Awarded in recognition of GPA in the top 15% of computer and information science juniors and seniors at Columbia
-
REU Travel Grant, National Science Foundation, 2023
Awarded $1200 to attend MobiHoc '23 and present my paper, Toward a Better Understanding of the Emotional Dynamics of Negotiation with Large Language Models.
-
NSF Research Experiences for Undergraduates Intern, University of Southern California, 2023
-
Distributed Research Experiences for Undergraduates Intern, Computing Research Association, 2022
Awarded $7000 to conduct research at Rice University with Professor Vicente Ordóñez Román.
-
Computer Science Research Mentorship Program Scholar, Google, 2021
Research Experience
-
Blablablab, Department of Electrical Engineering and Computer Science, University of Michigan, Ann Arbor, MI
Conduct research at the intersection of NLP, computational social science, and sociolinguistics.
-
Speech Lab, Department of Computer Science, Columbia University, New York, NY
Investigated relationship between empathy and code-switching by detecting empathetic utterances in code-switched speech, using RoBERTa fine-tuned for empathy detection. Investigated relationship between dialog acts and code-switching, by adapting dialog act annotation scheme to better suit, and annotating dialog acts on, code-switched speech. Wrote script to fine-tune Wav2Vec2 for end-to-end dialog act classification on Switchboard Dialog Act Corpus, using PyTorch and Hugging Face Transformers. Contributed to multimodal, multilingual empathetic speech corpus by processing video, audio, text with Jupyter Notebooks, Praat, FFmpeg.
-
Affective Computing Lab, Institute for Creative Technologies, University of Southern California, Los Angeles, CA
Developed negotiating agent for online studies of human negotiation behaviors. Engineered large language model prompts for agent NLU/NLG with OpenAI API. Built user interface using HTML, CSS, JavaScript, Flask. Assembled Qualtrics survey for crowdworkers to evaluate agent.
-
Vision, Language, and Learning Lab, Department of Computer Science, Rice University, Houston, TX
Quantified visual and linguistic complexity in MSCOCO image captioning dataset by creating and applying novel visual complexity metric, then fine-tuning BERT to predict visual complexity from image captions. Probed and mitigated content-related biases in resulting models.
-
Natural Language Text Processing Lab, Department of Computer Science, Columbia University, New York, NY
Generated semantic annotations for text generation/summarization in the literary domain.
-
Corter Lab, Department of Human Development, Columbia University, New York, NY
Identified communication strategies in collaborative learning dialogues by annotating speech acts and analyzing n-gram frequencies with R. Assisted with onboarding new research assistants. [Code]
Projects
-
COMS W4705: Natural Language Processing, Image Captioning System
Implemented image captioning system trained on Flickr 8k dataset, using Inception V3 (for feature extraction) and bidirectional LSTM.
-
LING GU4120: Language Documentation and Field Methods, Documenting the Sakha Language
Elucidated and documented the phonology, morphology, and syntax of the Sakha language by interviewing native speaker and analyzing collected transcriptions and audio recordings. [Report on Sakha Phonology] [Report on Sakha Nominal Morphology and Syntax] [Report on Sakha Verbal Morphology and Clausal Syntax] [Analysis of Two Sakha Texts]
-
COMS W4705: Natural Language Processing, Dependency Parser
Implemented neural network dependency parser using Keras API of TensorFlow and trained on WSJ subset of Penn Treebank corpus to achieve a macro average unlabeled attachment score of 76.17.
-
COMS W4701: Artificial Intelligence, American Sign Language Alphabet Recognition Model
Designed and implemented convolutional neural network using Keras API of TensorFlow to interpret images of American Sign Language letters with 92.2% accuracy.
Teaching Experience
-
Columbia COMS4771: Machine Learning, New York, NY
Introduction to machine learning. Held office hours, took attendance, monitored class discussion board, and graded assignments.
-
Columbia LING4903: Syntax, New York, NY
Introduction to modern formal theories of syntax. Held office hours, took attendance, and graded assignments.
-
Columbia COMS3251: Computational Linear Algebra, New York, NY
Introduction to linear algebra and its applications to data analysis. Hosted weekly office hours, moderated class discussion board, graded homework assignments, proctored final exam.
-
Columbia COMS1404: Emerging Scholars Program Seminar, New York, NY
Facilitated introductory computer science seminar to increase retention of diverse undergraduates. Moderated discussions, took attendance, and emailed weekly announcements.
Work Experience
-
Columbia Daily Spectator, New York, NY
Reported stories incorporating data analysis and visualization using Python, Adobe Illustrator, HTML, and CSS. Co-authored article on disparities in access to healthy food in local area, including interactive map of food sources. Provided data visualizations for article on COVID-19 vaccination rate disparities between different demographic groups. Recruited and trained reporters in data journalism. [Articles]
-
Columbia Daily Spectator, New York, NY
Reported story on student mental health in online communities, including communicating methodology and results of social media data analysis. [Articles]
-
Columbia Daily Spectator, New York, NY
Pitched, wrote, and published ~3,000-word feature in The Eye magazine, working on a two-week deadline. Conferred with editors, conducted background research, interviewed 6 sources, and wrote article. [Articles]
Science Communication
-
Columbia Science Review, New York, NY
Wrote biweekly articles on topics including environmental science, computational social science, biochemistry, and mathematics.
-
Columbia Science Review, New York, NY
Launched Instagram account (@csrcovidhub) to provide Columbia-relevant COVID-19 information to 200+ followers.
Service
-
Columbia University Women in Computer Science, New York, NY
Organized academic-related programming to support gender-diverse Columbia University computer science students, including workshops on course selection and lunches with female computer science professors. Hosted weekly coffee chats to provide students with advising on computer science internships, research, and courses.
-
Columbia University Application Development Initiative, New York, NY
Supported 2 Columbia University undergraduates majoring in computer science with one-on-one academic and preprofessional advising.
-
Columbia University Women in Computer Science, New York, NY
Organized activities focused on building a gender-diverse community of Columbia University computer science students, including an information session introducing the technical interview process, as well as social activities such as arts and crafts nights. Hosted weekly coffee chats to provide students with advising on computer science internships, research, and courses.