Avatar

Grace Lawley

PhD candidate, Computer Science & Engineering

Oregon Health & Science University

I am a Computer Science & Engineering PhD candidate at Oregon Health & Science University (OHSU) in Portland, Oregon, USA. My advisor is Steven Bedrick . My PhD research involves developing computational and statistical methods to detect and quantify different aspects of the language of children with Autism Spectrum Disorder (ASD).

I am also a Data Science Mentor at Posit (previously known as RStudio) for Posit Academy . I mentor small groups of working professionals as they develop and hone their data science skills in R or Python.

I recently presented two papers at the SIGDIAL 2023 conference in September: “A Statistical Approach for Quantifying Group Difference in Topic Distributions Using Clinical Discourse Samples” and “Computational Analysis of Backchannel Usage and Overlap Length in Autistic Children” .

**I am expecting to graduate in December 2023 and am currently on the job market!**

Interests

  • Computational Linguistics, Statistics
  • R, Python
  • Data Science, Data Visualization
  • Discrete Math, Statistics
  • Speech & Language Disorders

Education

  • PhD in Computer Science & Engineering, Expected Dec 2023

    Oregon Health & Science University

  • BA in Mathematics, 2017

    Lewis & Clark College

Publications

(2023). A Statistical Approach for Quantifying Group Difference in Topic Distributions Using Clinical Discourse Samples. Proceedings of the 24th Meeting of the Special Interest Group on Discourse and Dialogue (SIGDIAL).

PDF Code Slides

(2023). Computational Analysis of Backchannel Usage and Overlap Length in Autistic Children. Proceedings of the First Workshop on Connecting Multiple Disciplines to AI Techniques in Interaction-centric Autism Research and Diagnosis (ICARD 2023).

PDF Slides

(2022). Sex Differences in Pronoun and Maze Usage in the Language of Children wth Autism Spectrum Disorder. International Society for Autism Research Annual Meeting (INSAR).

Abstract Book Poster

(2022). "Um" and "Uh" Usage Patterns in Children with Autism: Associations with Measures of Structural and Pragmatic Language Ability. Journal of Autism and Developmental Disorders (JADD).

DOI

Talks

A Statistical Approach for Quantifying Group Difference in Topic Distributions Using Clinical Discourse Samples
Computational Analysis of Backchannel Usage and Overlap Length in Autistic Children

Posts

Multidimensional Scaling with Food Emoji 🥘🦐🦑

What is Multidimensional Scaling? Multidimensional Scaling (MDS) is a dimensionality reduction technique that is useful for exploratory data visualization. Some other popular dimensionality reduction techniques include Principle Component Analysis (PCA) and t-Distributed Stochastic Neighbor Embedding (t-SNE). Given a similarity matrix (e.g. a distance matrix), MDS projects the data points into an N-dimensional space while minimizing the amount of similarity information lost. In the ideal case, the closer the projected points are to one another, the more similar the are while the farther apart they are, the less similar they are.

Kuczaj Corpus Sentiment Analysis, pt 2

A continuation of a sentiment analysis of child language acquisition data using the Kuczaj Corpus from the CHILDES database.

Kuczaj Corpus Sentiment Analysis, pt 1

A sentiment analysis of child language acquisition data using the Kuczaj Corpus from the CHILDES database that I did for my final visualization project for CS 632.

Projects

*

Kuczaj Corpus Sentiment Analysis

Final project for CS 631 ‘Principles & Practice of Data Visualization’