Browsing by Author "Brown, Daniel"

Now showing 1 - 6 of 6

Assessment of AI-Generated Pediatric Rehabilitation SOAP-Note Quality
(University of Waterloo, 2025-02-19) Amenyo, Solomon; Grossman, Maura; Brown, Daniel; Wylie-Toal, Brendan
This study explores the integration of artificial intelligence (AI) or large language models (LLMs) into pediatric rehabilitation clinical documentation, focusing on the generation of SOAP (Subjective, Objective, Assessment, Plan) notes, which are essential for patient care. Creating complex documentation is time-consuming in pediatric settings. We evaluate the effectiveness of two AI tools; Copilot, a commercial LLM, and KAUWbot, a fine-tuned LLM developed for KidsAbility Centre for Child Development (an Ontario pediatric rehabilitation facility), in simplifying and automating this process. We focus on two key questions: (i) How does the quality of AI-generated SOAP notes based on short clinician summaries compare to human-authored notes, and (ii) To what extent is human editing necessary for improving AI-generated SOAP notes? We found no evidence of prior work assessing the quality of AI-generated clinical notes in pediatric rehabilitation. We used a sample of 432 SOAP notes, evenly divided among human-authored, Copilot-generated, and KAUWbot-generated notes. We employ a blind evaluation by experienced clinicians based on a custom rubric. Statistical analysis is conducted to assess the quality of the notes and the impact of human editing. The results suggest that AI tools such as KAUWbot and Copilot can generate SOAP notes with quality comparable to those authored by humans. We highlight the potential for combining AI with human expertise to enhance clinical documentation and offer insights for the future integration of AI into pediatric rehabilitation practice and other settings for the management of clinical conditions.
Modelling Chart Trajectories using Song Features
(University of Waterloo, 2019-08-23) Perrie, Jonathan; van Beek, Peter; Brown, Daniel
Over the years, hit song science has been a controversial topic within music information retrieval. Researchers have debated whether an unbiased dataset can be constructed to model song performance in a meaningful way. Often, classes for modelling are derived from one dimension of song performance, like for example, a song’s peak position on some chart. We aim to develop target variables for modelling song performance as trajectory patterns that consider both a song's lasting power and its listener reach. We model our target variables over various datasets using a wide array of features across different domains, which include metadata, audio, and lyric features. We found that the metadata features, which act as baseline song attributes, oftentimes had the most power in distinguishing our proposed task classes. When modelling hits and flops along one dimension of song success, we observed that the dimensions carried contrasting information, thus justifying their fusion into a two-dimensional target variable, which could be useful for future researchers who want to better understand the relationships between song features and performance. We were unable to show that our target variables were all that useful for modelling more than two classes, but we believe that this is more a limitation of the features, which were often high level, rather than the target variables' separability. Along with our model analysis, we also carried out a re-implementation of a related study by Askin & Mauskapf and considered different applications of our data using methods from time series analysis.
People know how diverse their music recommendations should be; why don’t we?
(University of Waterloo, 2021-02-17) Robinson, Kyle; Brown, Daniel
While many researchers have proposed various ways of quantifying recommendation list diversity, these approaches have had little input from users on their own perceptions and preferences in seeking diversity. Through a set of user studies we provide a better understanding of how users view the concept of diversity in music recommendations, and how intra-list diversity can be adapted to better represent their diversity preference. Our results show that users have a clear idea of what music recommendation diversity means to them, accuracy metrics do not model overall list satisfaction, and filtering recommendations on genre before list diversification can positively impact list satisfaction. More importantly, our results highlight the need to base music recommendation metrics on insights from real people
Style Recognition in Music with Context Free Grammars and Kolmogorov Complexity
(University of Waterloo, 2020-03-11) Mondol, Tiasa; Brown, Daniel
The Kolmogorov Complexity of an object is incomputable. But built in its structure is a way to specify description methods of an object that is computable in some sense. Such a description method then can be exploited to quantify the bits of information needed to generate the object from scratch. We show that Context-Free Grammars form such a viable description method to specify an object and the size of the grammar can be used to estimate the Kolmogorov Complexity. We use such estimation in approximating the Information Distance between two musical strings. We also show that such distance measure in music can be used to recognize the genre, composer and style and also for music classification.
Technology Design Recommendations Informed by Observations of Videos of Popular Musicians Teaching and Learning Songs by Ear
(University of Waterloo, 2024-07-11) Liscio, Christopher; Brown, Daniel
Instrumentalists who play popular music often learn songs by ear, using recordings in lieu of sheet music or tablature. This practice was made possible by technology that allows musicians to control playback events. Until now, researchers have not studied the human-recording interactions of musicians attempting to learn pop songs by ear. Through a pair of studies analyzing the content of online videos from YouTube, we generate hypotheses and seek a better understanding of by-ear learning from a recording. Combined with results from neuroscience studies of tonal working memory and aural imagery, our findings reveal a model of by-ear learning that highlights note-finding as a core activity. Using what we learned, we discuss opportunities for designers to create a set of novel human-recording interactions, and to provide assistive technology for those who lack the baseline skills to engage in the foundational note-finding activity.
TwitSong: A current events computer poet and the thorny problem of assessment.
(University of Waterloo, 2018-11-29) Lamb, Carolyn; Brown, Daniel; Clarke, Charles
This thesis is driven by the question of how computers can generate poetry, and how that poetry can be evaluated. We survey existing work on computer-generated poetry and interdisciplinary work on how to evaluate this type of computer-generated creative product. We perform experiments illuminating issues in evaluation which are specific to poetry. Finally, we produce and evaluate three versions of our own generative poetry system, TwitSong, which generates poetry based on the news, evaluates the desired qualities of the lines that it chooses, and, in its final form, can make targeted and goal-directed edits to its own work. While TwitSong does not turn out to produce poetry comparable to that of a human, it represents an advancement on the state of the art in its genre of computer-generated poetry, particularly in its ability to edit for qualities like topicality and emotion.