EILEEN BLUM

I am a seventh year abd PhD student in the linguistics department at Rutgers University. I am interested in computational phonology and the phonetics-phonology interface. My first research project investigated the acoustic properties of word stress in Munster Irish. My dissertation uses a formal language theory approach to investigate the computational complexity of vowel harmony patterns over multi-tiered autosegmental representations. I plan to complete my dissertation in the summer of 2022 for an October dated degree.

Outside of academia and work, I also enjoy training animals and medieval armored combat. I have ridden and trained off-the-track thoroughbreds in the hunter/jumper discipline for over ten years. I trained my first dog as a kid and now I am training my cat to perform some basic tasks on cue. I also participated in heavy combat in the Society for Creative Anachronism (SCA) and I enjoy larp and boffer fighting with an affinity for pole weapons, particularly glaive and axe.

EDUCATION

PHD LINGUISTICS, RUTGERS UNIVERSITY

September 2015 - October 2022

Google Cloud CCAI

• Annotate conversation data to identify voice and chat bot failures and successes

• Identify data quality issues, Create a plan, and Resolve NLU and UX problems; Implement changes directly within DialogflowCX

• Collaborate with team members to accomplish complex tasks and with engineers to improve processes

• Serve as onboarding ambassador to support new team members

• Write task and process documentation to track changes and train new team members

• Write best practices documentation for Dialogflow CX, published by Google

• Proofread and edit documentation for SCRAPI Python library

Research: Phonology, Formal Language Theory, Computational complexity

• Executed two major research projects: design, structure, analysis, and maintained schedule from inception to completion

• Develop theory to explain how abstract representations affect the computational complexity of sound patterns across languages

• Recorded 20 native speakers to determine word stress pattern in Munster Irish

• Collaborate with experts in four disciplines: computer science, math, linguistics, and cognitive psychology

• Strong proficiency with IPA and excellent understanding of other phonological representations

• Taught two introduction to linguistics courses with 30 students each, Coordinate recitation section of 15 students that reinforced lecture content

• Created course content and served as primary instructor for two expository writing courses with up to 30 students each; Wrote response memos to student drafts detailing plans for improvement and revisions

• Organized and hosted first PhD to Industry informational event with five panelists and up to 50 attendees

• Orchestrated summer mini-course with five lessons on methods of artificial learning

• Coordinated colloquium series for two years

I apply formal language theory to natural language data in order to analyze the computational complexity of vowel harmony patterns across both well studied and understudied languages. I use this computational approach to investigate the effects of different representational data structures on complexity and I develop a new theory of autosegmental locality.

My partner and I built a Wasserstein Generative Adversarial Network (WGAN) in Python to generate automated song lyrics in lines of 8 words at a time. We compared our WGAN with a Soft-GAN trained on the same dataset. We trained both GANs on a Kaggle dataset of metal song lyrics, which we processed using NLTK and pandas. The GANs were built using Keras, Tensorflow, and Numpy. Lastly, we calculated BLEU scores for both models and determined that neither generated very natural sounding lyrics: WGAN received all 0s, Soft-GAN averaged 0.06 for n-grams of length 1-4.

My partner and I created a classifier in Python to distinguish song lyrics by genre. We used two Kaggle data sets of song lyrics which were cleaned using the GenSim and NLTK packages. We then used the shallow neural network in the Word2Vec package to create high-dimensional word vectors, PCA and k-clustering to group them based on semantic similarity, and trained a DecisionTreeClassifier to distinguish lyric sets. The classifier achieved 81% accuracy.