PELDA: Platform for Endangered Language Documentation and Analysis
The Issues: Currently, about 90% of the world’s languages are endangered and many are quickly vanishing. Language Endangerment is considered one of the most urgent problems in humanities. While researchers have obtained massive amounts of digital recordings during fieldwork, they are in urgent need of computational support to annotate, analyze and manage such large volumes of data. To address this need, our long-term project goal is to architect and create a cloud-based computation and collaboration platform, namely PELDA (Platform for Endangered Language Documentation and Analysis), to enhance linguistics work on indigenous languages. As an important component, a prototype audio search system has been partially deployed based on our preliminary work supported by the NEH Digital Humanities Start-up Grant and Microsoft Azure Research Award.
Student Outcomes: Release researchers from the burden of maintaining and managing data files and software tools so they can focus on linguistic research activities. Enables effective collaboration from various aspects ranging from data file management, and working status tracking, to group communication. Enhances information query and retrieval mechanisms with multimedia data mining techniques to match the “multimedia” nature.
Student Qualifications: Fluent with Java (CSS162 completed) or C++ programming (CSS342 completed), Commit to the project for two quarters and Must register for CSS497 or CSS499 credits during the 1st or at the latest the 2nd quarter
Student Responsibilities: Use RESTful web services and cloud computing techniques to deploy existing linguistics tools to PELDA. Develop websites for research and dissemination of project materials. Enable version control and central repository capability to facilitate collaboration and status tracking. Improve audio processing and retrieval component.2 quarters minimum
Resources and additional content: Check out the prototype audio search system and version control component at http://www.pelda.us/