Celebrating Variation in Multilingual Contexts: Insights from NWAV Conference
December 13, 2024
Last month, my fellow computational linguist Claire O’Neill and I discussed our research at the New Ways of Analyzing Variation (NWAV) conference in Miami. This year’s conference, hosted by Florida International University and the University of Miami, celebrated ‘variation in multilingual contexts’ —a theme that reflects SoapBox’s and Curriculum Associates‘ commitment to embracing diversity and fostering inclusion.
Collecting Representative Voice Data
Our poster session presented our ongoing work in collecting representative voice data from Spanish-speaking children across the US. We worked to deliver a diverse dataset that would represent a wide range of dialects, backgrounds, ages, and types of speech, including both read and spontaneous speech. This diversity is crucial for accurately assessing our Spanish voice engine’s performance for Spanish-speaking communities and in real-world educational scenarios.
Key Takeaways
Recognition for Our Approach
Our thoughtful and inclusive method of collecting diverse voice data resonated strongly with attendees, affirming the importance of representation in building equitable voice solutions.
Excitement About Classroom Applications
There was significant interest in how our voice technology could transform learning environments, supporting teachers’ great work and improving student outcomes.
Unique Dataset
Our dataset was notable as unique in its size and scope and offers a valuable resource for investigating the language behaviors of children.
Giving Back to Communities
A recurring theme at the conference was the importance of giving back to the communities from whom we analyze data. While research is driven by the pursuit of knowledge, our work must benefit those who give up their time and lend their voices to such projects.
We are fortunate that our research shapes the development of a voice engine designed to power learning tools teachers and students use. We can apply our findings to improve our voice engine and provide valuable and accurate assistance to educators.
Shaping the Future of Linguistic Research
Language is inherently social—you can’t study it without considering the people using it. Similarly, you can’t separate the product from the users in the language technology space. It’s vital we continue to center students first and foremost as we build and develop our Spanish voice engine.
It’s vital we continue to center students first and foremost as we build and develop our Spanish voice engine.
Language is constantly changing and evolving; it is not static, nor should our work be. By staying up to date with new developments, tools, and analyses, we can continuously provide the best possible experience for teachers and their students.
SoapBox Labs is now part of Curriculum Associates, a leader in K-12 instruction and assessment solutions. To learn more about i-Ready and other award-winning CA products, go to www.cainc.com
Author: Emma O’Neill, Computational Linguist & NLP Engineer, SoapBox Labs