Three students present at spring Undergraduate Research Symposium
Using Meta-Learning to address the task-agnostic problem in Natural Language Understanding
Student: Tianyi Sun
Major: Mathematics
Mentor: Maria Gini
Abstract: Human beings can generalize from a single example of a task to similar tasks. However, computers need to effectively see thousands of examples to learn to generalize to similar tasks. To learn computers need to be given labeled datasets, which include both correct and incorrect examples, properly labeled. The limited amount of available labeled datasets makes it difficult for machine learning to achieve human-level performance. Meta-Learning, which is known as "learning to learn", can address this difficulty. This means that Meta-Learning enables learning with a few training samples and enables adaptation across domains or in a single constantly changing domain. Meta-Learning has been explored and used in robotics and computer vision, but it has not been sufficiently explored in natural language processing. This research aims to explore how to use Meta-Learning in addressing task-agnostic problems in Natural Language Understanding.
Porting Mobile App Functionality across Platforms
Student: Zhaoqing (Andrew) Li
Major: Computer Science
Mentor: Mattia Fazzini
Abstract: Google is developing a new operating system (OS) for mobile apps called Fuchsia and this OS is the candidate for replacing the Android OS in the near future. To take full advantage of Fuchsia, Android developers will need to migrate their apps to the new OS. One way to migrate existing apps to the new OS is to migrate their apps so that they can be developed through the Flutter framework. To perform the migration task, developers need to re-design and re-implement portions of their apps.
In this project, we investigated how developers migrate their apps as a first step in building automated tools for aiding this time-consuming and error-prone task. Specifically, we focused on comparing how GUI is written for native Android applications and Flutter applications and if the migration requires implementing new methods. We selected three github repositories which are transforming their applications from native android to flutter based on their version control history, and analyzed the elements in the android layout files before the transformation, and corresponding dart files after migrating to flutter. We then find the matching widgets in flutter for each layout file from the native codebase, build, run and check if the two versions of the application have the same GUI elements, and compare the underlying logic from native activity files and dart classes to understand how developers change their app logic. We concluded the study by reporting statistics related to the migration task.