SRC Grand Finalists 2023

GRADUATE CATEGORY

First Place:
Zhe Liu - Chinese Academy of Sciences
ICSE 2022
Title of Submission: Woodpecker: Identifying and Fixing UI Display Issues in Mobile Application

Second Place:
Juan Carlos Alonso Valenzuela - University of Seville
ESEC/FSE 2022
Title of Submission: Automated Generation of Test Oracles for REST APIs

Third Place:
Irene Zanardi - Università della Svizzera Italiana, Switzerland
CHI 2022
Title of Submission: Design and Assessment of Hoomie, a Small Multisensory Space for Autistic Children in Primary Schools

UNDERGRADUATE CATEGORY

First Place:
Takahito Murakami - University of Tsukuba
SIGGRAPH 2022
Title of Submission: Kuchibashi: 3D-Printed Tweezers Bioinspired by the New Caledonian Crow’s Beak

Second Place:
Raphael Douglas Giles - University of New South Wales
SPLASH 2022
Title of Submission: Termination of Recursive Functions by Lexicographic Orders of Linear Combinations

Third Place:
Christopher Bain - University of Maryland, Baltimore County
TAPIA 2022
Title of Submission: Using Supervised Machine Learning Methods to Create a Gene-Based ALS Predictor from Postmortem Transcriptomics Data

ACM Student Research Competition

The ACM Student Research Competition is an internationally recognized venue enabling undergraduate and graduate students to experience the research world, share research results and exchange ideas, rub shoulders with academic and industry luminaries, understand the practical applications of their research and gain recognition.

2023 SRC Winner, 1st Place, Graduate Category

Zhe Liu, Chinese Academy of Sciences

"Woodpecker: Identifying and Fixing UI Display Issues in Mobile Application" (ICSE 2022)

The complexity of GUI and some combinations of personalized settings make the UI display issues occur frequently. Unfortunately, little is known about the causes of UI display issues. The Android fragmentation and variety of UI components post a great challenge
to repair the issue. Based on the our empirical study, this paper proposes Woodpecker to automatically detect, localize and repair UI display issues in Android apps. It detects screenshots with issues with computer vision technology, localizes buggy source code from
screenshot, and repairs issues with pre-defined templates automatically. We evaluate Woodpecker with 30 real-world UI display issues, it can successfully detect 87% and repair 91% issues. We further apply Woodpecker to another 316 popular open-source Android apps,
and successfully uncover 123 previously-undetected issues. It can automatically repair 116 (94%) issues, with 106 of them accepted by developers so far, while others pending (none of them is rejected).
The Woodpecker demo video link: https://youtu.be/uPXEtqaIP8g.[Read more]

2023 SRC Winner, 2nd Place, Graduate Category

Juan Carlos Alonso Valenzuela, University of Seville

"Automated Generation of Test Oracles for REST APIs" (ESEC/FSE 2022)

Test case generation tools for REST APIs have grown in number and complexity in recent years. However, their advanced capabilities for automated input generation contrast with the simplicity of their test oracles (i.e., mechanisms for determining whether a test has passed or failed), which limits the types of failures they can detect to crashes, regressions, and violations of the API specification or design best practices. In this paper, we present AGORA, an approach for the automated generation of test oracles for REST APIs through the detection of invariants—properties of the output that should always hold. In practice, AGORA aims to learn the expected behavior of an API by analyzing previous API requests and their corresponding responses. For this, we extended the Daikon tool for dynamic detection of likely invariants, including the definition of new types of invariants and the implementation of an instrumenter called Beet. [Read more]

2023 SRC Winner, 3rd Place, Graduate Category

Irene Zanardi, Università della Svizzera Italiana, Switzerland

"Design and Assessment of Hoomie, a Small Multisensory Space for Autistic Children in Primary Schools" (CHI 2022)

ABSTRACT While Multisensory Environments (MSEs) are thought to be useful in addressing autistic children’s sensory processing dysfunction, there is a lack of solutions tailored to schools that address students’ anxiety while meeting the needs of support teachers. We describe the iterative design process that resulted in the creation of Hoomie, a small relaxing MSE, as well as the 2-week exploratory study that involved six classes in a primary school. We present the findings of the study alongside the overall contribution of the research project.

ACM Reference Format:
Irene Zanardi. 2023. CHI: G: Design and Assessment of Hoomie, a Small
Multisensory Space for Autistic Children in Primary Schools. In Proceedings
of ACM Conference (Conference’17). ACM, New York, NY, USA, 5 pages.
https://doi.org/10.1145/nnnnnnn.nnnnnnn

[Read more]

2023 SRC Winner, 1st Place, Undergraduate Category

Takahito Murakami, University of Tsukuba

"Kuchibashi: 3D-Printed Tweezers Bioinspired by the New Caledonian Crow’s Beak" (SIGGRAPH 2022)

This study explores the ergonomic potential of tweezers and presents the biomimicry-inspired Kuchibashi prototype, based on the New Caledonian Crow’s beak. Balancing aesthetics and functionality, the Kuchibashi improves graspability and psychological friendliness. A user study comparing the Kuchibashi to traditional tweezers and fingers demonstrates competitive performance with smaller beads and excellent performance with larger beads, with users appreciating its sense of security, ease of grasp, and attractive design. Despite limitations, the study highlights the untapped potential of biomimetic design in human-tool interaction. The versatile Kuchibashi shows promise as a functional and aesthetically pleasing tool for applications in tableware, healthcare, and manufacturing. CCS Concepts: • Applied computing→Arts and humanities; • Human-centered computing→Interaction design. Additional Key Words and Phrases: 3D printing, New Caledonian Crow, ergonomics , human-tool interaction. [Read more]

2023 SRC Winner, 2nd Place, Undergraduate Category

Raphael Douglas Giles, University of New South Wales

"Termination of Recursive Functions by Lexicographic Orders of Linear Combinations" (SPLASH 2022)

1 Problem and Motivation The widely used proof assistant Isabelle/HOL has a simple yet effective algorithm [3] for generating proofs that certain recursive functions terminate. Isabelle/HOL’s termination checker works by finding a lexicographic ordering on the measures generated by each argument of the function usingthe structural ordering [2] which decreases at every recursive call. This project seeks to extend this algorithm to check for linear combinations of these measures with coefficients in the natural numbers to prove that a broader class of functions terminate. 2 Background and Related Work 2.1 Preliminaries In functional languages, non-termination comes from one source: recursion. A function which does not recurse, and which only calls functions that do not recurse, will assuredly terminate. Thus, to determine when a function terminates, we must inspect the recursive structure of that function. Recursive functions may be defined by a collection of pattern matching rules, similar to Haskell. Each line in a function definition defines a pattern for the input data to match, and an output term for the function to return in that case. We call these the defining equations of that function. For example, the binomial coefficient can be written in this form (Figure 1). choose (0, 𝑘) = 1 choose (𝑛, 0) = 1 choose (S 𝑛, S 𝑘) = choose (𝑛, S 𝑘) + choose (𝑛, 𝑘) [Read more]

2023 SRC Winner, 3rd Place, Undergraduate Category

Christopher Bain, University of Maryland, Baltimore County

"Using Supervised Machine Learning Methods to Create a Gene-Based ALS Predictor from Postmortem Transcriptomics Data" (TAPIA 2022)

I. ABSTRACT Amyotrophic Lateral Sclerosis (ALS) is a devastating and progressive neurological disorder that results from the degeneration of motor neurons in the brain and spinal cord. Despite the identification of several pathogenic mutations, the majority of ALS cases are sporadic and lack a clear genetic cause. In this study, we utilized supervised machine learning techniques to analyze transcriptomic data from ALS post-mortem motor cortex samples to identify gene expression patterns that differentiate ALS from healthy controls. Using a logistic regression classifier on a dataset of 276 post-mortem transcriptomic samples collected from multiple academic institutions, we were able to develop a model that predicted ALS with a high degree of accuracy, achieving an F1 score of 90.9%. We also performed decision weight rankings to identify genes that were most strongly associated with the ALS samples, and found that our model accurately identified genes that are currently the subject of active research. Overall, our findings suggest that machine learning techniques can be used to develop accurate predictive models for ALS postmortem tissue, and that these models have the potential to help generate hypotheses for biological pathways that contribute to ALS. [Read more]