CAPSTONE PROJECTS

Find out more about our Capstone projects

Natural Language Processing (NLP) is a rapidly growing field with applications in many of the technologies we are all accustomed to using every day, from virtual assistants and smart speakers to autocorrect functions.

Our Masters program at UCSC balances theory with practice including a 15 unit Capstone project to enable you to apply the skills you’ve acquired on the program to a real-world issue or challenge. You’ll gain valuable employability skills and practical experience both from working in a team with your peers and from the insights you’ll gain from your academic or organizational mentor(s). Recent student teams have collaborated with experts from industry giants like Bloomberg, IBM, Interactions, LinkedIn, Microsoft and others to address industry-relevant and research-focused topics in the NLP field.

The Capstone project is a great way for you to scope out the NLP field, looking at projects that interest you and identifying organizations that you may wish to work with in future. Working closely with your mentor during the project you’ll extend your networks and gain practical employability skills that will support your future job search. Many organizations value their involvement in the Capstone projects as this can be a useful way to informally evaluate students’ skills and find a good fit for workplace vacancies.

The NLP Capstone experience offers a great opportunity to extend your networks and put yourself in front of potential employers. Registration is open to the professional NLP community and invitations are sent to a range of interested parties who are able to attend either in person or remotely.

Since our program launched in 2020 our NLP students have collaborated with industry partners at Adobe, Amazon, Bloomberg, CDIAL, IBM, Intel, Interactions, Google, LinkedIn, Meta, Microsoft, ModelCode and UCSC. Scroll down to see the projects, and follow us on LinkedIn for updates from our Capstone teams.

2024 Capstone Projects

Summarization for Long Document Input using Multiple LLMs

Alex Liu, Ethan Liu, June Kim, Michael Fang, Nikhil Singh, Yash Bhedaru
Adobe (Hanieh Deilamsalehy)

GPS4LLM: Graph-based Planning System for Large Language Models

Anish Pahilajani, Devasha Trivedi, Jincen Shuai, Khin Yone, Neng Wan, Samyak Rajesh Jain
Meta (Namyoun Park)

Personalized Graph-based Retrieval for Large Language Models

Cameron Dimacali, Ojasmitha Pedirappagari, Steven Au
Intel (Nesreen Ahmed)

GUT-Bench: Maintainability and Performance

Decker Krogh, Ethan Sin, Esha Ubale, Sreekar Molakalapalli, Sujit Noronha
ModelCode (Antoine Raux)

2023 Capstone Projects

Open Domain Multimodal QA

Brian Farrell, Apala Thakur and Sugam Garg

Meta (Bing Liu)

Model Interpretability For Hallucination Detection In Summaries

Dhananjay Sonavane, Bhrigu Garg and Devavrat Joshi, Dhananjay Sonavane, Meenal Chavan, Parsa Mazaheri and Priyesh Vakharia
UCSC (Leilani Gilpin & Ian Lane)

Dialogue Act Labelling for Open Domain Dialogue Systems

Haoran Zhang, Nan Qiang, Neha Pullabhotla,
Sidharth Babu and Xin Zhang
UCSC (Angela Ramirez & Marilyn Walker)

Adapters and Reinforcement Learning for Data-Efficient Machine Translation

Abigail Kufeldt, Malini Kar, Pranjali Basmatkar, Kushagra Seth
Shashwat Pandey, Vignesh S and Parikshith Honnegowda
CDIAL (Olayinka Iyinolakan)

Temperature Guided Text-to-Image Generation with Reinforcement Learning

Cookie Pan, Haolong Jia, Zoe Gupta, Mridul Pankaj Khanna,
Vijay Chilaka, Chengxuan Xia and Sree Latha
Adobe (Ryan Rossi)

2022 Capstone Projects

Exploring Code Style Transfer with Neural Networks

IBM Research

Neural Models of Supertagging for Semantic Role Labeling and Beyond

Interactions

Comparing Dictionaries and Word Embeddings

LinkedIn

Multimodal Knowledge Extraction and Question Answering in Farming

UCSC and Google X

2021 Capstone Projects

Amazon Alexa Prize Discourse Model & Coreference Resolution

UCSC

Identifying Errors in SRL using Weak Supervision

IBM

Domain Adaptation for Question Answering on COVID-19

Amazon and IBM

MKD-Ultra: Compressing Causal Language Models in Multiple Steps

Microsoft

Informers: Evaluating Explanation Quality

IBM Research

Information Extraction of Corporate Events from the Web

Bloomberg

ABOUT US

PROGRAM

FINANCIALS

APPLY NOW

WORKING WITH INDUSTRY

YOUR CAREER

HUMANS OF NLP

THE SILICON VALLEY CAMPUS

CONTACT

ADMISSIONS

CAPSTONE

PEOPLE