Language Technologies Institute
Capstone Programs
Master of Computational Data Science (MCDS)
January–December
Sponsorship Fee: $40,000 (using public data); $80,000 (using private data)
The MCDS program trains professional master’s students in all aspects of design, engineering and deployment of very large information systems. The multidisciplinary MCDS brand of computational data science rests on three pillars, namely analytics, systems and human-computer interaction. MCDS students gain a solid foundation in all three of these while specializing in one they choose as their concentration. In particular, students in this program learn to build solutions for complex, noisy and incomplete data sources using cutting-edge AI and machine learning in order to make the full gamut of multimodal data sources do valuable work in ways that are innovative, ethical, safe, efficient, effective and able to benefit both societal good and personal quality of life. They learn to do this work with web-scale storage and computing systems and incorporate human-centered designs specific to the developers and users of the analytics and results.
Capstone projects leverage all of this potential to develop solutions that challenge the bounds of the state-of-the-art and meet the real-world needs of clients, with teams potentially including students from all three concentrations. Housed in the Language Technologies Institute, MCDS draws from faculty members throughout SCS, including the Computer Science Department, Machine Learning Department and the Human-Computer Interaction Institute.
Program Lead
Carolyn Rosé
Professor and Director
Master of Computational Data Science Program
Master of Science in Intelligent Information Systems (MIIS)
January–December
Sponsorship Fee: $40,000 per team, teams of 2-4 students
MIIS is a practice-oriented, professional degree program that offers both advanced study and practical experience in the processing and analysis of unstructured and semistructured information (such as text, image, video, speech and audio); mining; and intelligent information technologies.
Past Projects Include
- Abstractive Summarization of Medical Conversations.
- Avoiding Derailment in Online Conversations.
- Code-Mixed (Multilanguage) Language Understanding.
- Conversational Assisted Search.
- Customer Service Call Analysis.
- Describing Similarities and Differences Between Video Segments.
- Intelligent Code Completion.
- Predicting Patients’ Conversation Transitions in Online Health Support Groups.
- Recipe Generation From Ingredients.
- Smart Reply for Text Messages.
- Visual Question Generation From Video Clips.
Program Leads
Master of Science in Artificial Intelligence and Innovation (MSAII)
September–May (fall and spring semesters)
Sponsorship Fee: $25,000 for a small entity (<300 employees); $50,000 for a large entity (300+ employees)
The M.S. in Artificial Intelligence and Innovation Program (MSAII) is a four-semester professional master’s degree offered by the Language Technologies Institute. The first two semesters are spent on the building blocks of AI: mathematical foundations, machine learning, natural language processing and engineering AI systems. Students work at company internships during the summer semester.
In the last two semesters of the program, students plan and complete a capstone project. During the fall, they take AI Innovation, a course in which problems posed by sponsors are selected by teams of six or seven students, who then propose an innovative AI application to solve the problem. During the course, the teams work with the sponsor to understand the business environment in which the problem has arisen, the interests of the stakeholders and the financial implications of the problem.
Each team works on a different project. The project is shaped during the AI Innovation course, which ends with a presentation to the sponsor of a proposed minimum viable product.
In the spring term, the same students develop a working AI system as part of the capstone course which, at 36 units, is equivalent to three full-semester courses. The resulting system is licensed to the sponsor under the CMU Intellectual Property Policy.
The number of students entering second year in 2023 will be 37, so we will need six projects. The problem must involve an innovative AI application and should be feasible to implement within 3 person-years over the period October 2023 to May 2024. The teams will devote approximately 0.5 person-years to understanding the business case and defining the solution in the fall term, and approximately 2.5 person-years during the spring term completing the project.
After the projects are completed in May, all teams participate in a Capstone Prize Competition. A panel of independent judges evaluates presentations made by the teams and a substantial cash prize is given to each member of the winning team: the first prize is $2,000 per student, second prize is $1,000 per student, and third prize $500 per student.
Descriptions of the projects completed in 2022-23 can be found in this Google Drive.