Paul Bodily, PhD

Assistant Professor, Computer Science

Research Interests

Machine learning and computational creativity as it relates to addressing issues related to mental health and well-being.

Sample Publications

M. S. Fujimoto, C. A. Lyman, P. M. Bodily, M. J. Clement, and Q. Snell, “GNUMAP 4.0: Space and time efficient NGS read mapping using the FM-index,” Insights of Bioinformatics, vol. 1, no. 1, pp. 1-8, 2019.

Suvorov, N. Jensen, C. Sharkey, M. S. Fujimoto, P. M. Bodily, H. Wightman, T.Ogden, M. J. Clement, and S. M. Bybee, “Opsins have evolved under the permanent heterozygote model: insights from phylotranscriptomics of Odonata,” Molecular Ecology, vol. 26, no. 5, pp. 1306-1322, 2017.

P. M. Bodily, M. S. Fujimoto, J. T. Page, M. J. Clement, M. T. Ebbert, and P. G. Ridge, “A novel approach for multi-SNP GWAS and its application in Alzheimer’s disease,” BMC Bioinformatics, vol. 17, no. 7, pp. 455-463, 2016.

P. M. Bodily, M. S. Fujimoto, Q. Snell, D. Ventura, and M. J. Clement, “ScaffoldScaffolder: Solving contig orientation via bidirected to directed graph reduction,” Bioinformatics, vol. 32, no. 1, pp. 17-24, 2016.

P. M. Bodily, M. Fujimoto, C. Ortega, N. Okuda, J. C. Price, M. J. Clement, and Q. Snell, “Heterozygous genome assembly via binary classification of homologous sequence,” BMC Bioinformatics, vol. 16, no. 7, 2015

Bio

Professor Bodily’s research addresses the question of whether or not computers can exhibit autonomous creativity, focusing particularly on the domain of lyrical music composition and the challenge of global structure. His approach incorporates a modular machine learning framework called hierarchical Bayesian program learning, which facilitates breaking the problem of music composition into smaller pieces, and focuses primarily on developing machine learning models that solve the problems related to structure.

Of note, Dr. Bodily developed an adaptation of non-homogeneous Markov models that enables long-range constraints and a structural learning model adapted from the Smith-Waterman alignment method, which extends sequence alignment techniques from bioinformatics. He has incorporated these advances into a full-fledged computational creative system called Pop* (pronounced pop star) and has shown through various evaluative methods that the system can be argued to possess, to varying degrees, the characteristics of creativity. Professor Bodily's prior research also includes several bioinformatics publications on heterozygous genome assembly algorithms.

Dr. Bodily holds a Ph.D. and M.S. in Computer Science from Brigham Young University (BYU). He graduated summa cum laude from BYU with a B.S. in Bioinformatics, a B.A. in Italian, and minors in Computer Science and Music.