Phonological computation
General
- Undergraduate; 2nd year
- 10 learners; teams of 2
- 20 hours per learner
- Dates set by experience
- Learners self-assign
Preferred companies
- 1 projects wanted
- Anywhere
- Academic experience
- Large enterprise
- Any
Categories
Project timeline
-
March 8, 2022Experience start
-
April 3, 2022Experience end
Overview
- Details
-
Comparative assessment of different frameworks for the analysis of phonological patterns in natural language
- Learner skills
- Data analysis, Research, Problem analysis
- Deliverables
-
Weighted finite-state transducers (WFSTs) are widely used by engineers and
computational linguists for generating and processing text and speech. Using the methods presented by Gorman and Sproat (Finite-State Text Processing, 2021) students will create case studies that include the compilation and application of context-dependent
rewrite rules, the construction of morphological analyzers and generators, and text
generation and processing applications. Kyle Gorman (CUNY/Google) will guest lecture in the course.
Project Examples
Finite State Transducer implementations of the phonology of various natural languages
Additional company criteria
Companies must answer the following questions to submit a match request to this experience:
A representative of the company will be available to answer questions from students in a timely manner for the duration of the project.
A representative of the company will be available for a pre-selection discussion with the administrator of the course to review the project scope.