Probabilistic modeling with human factors in machine learning
| dc.contributor.advisor | Lease, Matthew A. | |
| dc.contributor.committeeMember | Wallace, Byron | |
| dc.contributor.committeeMember | Durrett, Greg | |
| dc.contributor.committeeMember | Liu, Qiang | |
| dc.creator | Nguyen, An Thanh | |
| dc.creator.orcid | 0000-0002-8990-7367 | |
| dc.date.accessioned | 2020-10-21T15:14:28Z | |
| dc.date.available | 2020-10-21T15:14:28Z | |
| dc.date.created | 2020-08 | |
| dc.date.issued | 2020-09-03 | |
| dc.date.submitted | August 2020 | |
| dc.date.updated | 2020-10-21T15:14:29Z | |
| dc.description.abstract | Although machine learning has been a very popular research area, the human factors have been largely unexplored. In this dissertation, we present our work in three directions: (1) models for better understanding human annotators, (2) systems for interacting with human users, and (3) software tools for human developers. Together, our models, systems, and tools aim at understanding and improving the interactions between humans and machine learning using a probabilistic approach. In the first direction, we present our work in probabilistic models for evaluating annotators, identifying patterns of annotation errors, and predicting subjective annotations. As for the second direction, we study user interaction in the task of fact-checking: predicting the veracity of a claim given reporting articles. We propose a probabilistic model that combines annotator accuracies, article stances, source reputation, and claim veracities. We also present the results of our user studies on how people interact with our system. In the third direction, we introduce our software tools for developing transparent machine learning systems. The tools integrate back-end machine learning models and front-end user interfaces, enabling developers to address the accuracy-transparency trade-off. | |
| dc.description.department | Computer Science | |
| dc.format.mimetype | application/pdf | |
| dc.identifier.uri | https://hdl.handle.net/2152/83387 | |
| dc.identifier.uri | http://dx.doi.org/10.26153/tsw/10384 | |
| dc.language.iso | en | |
| dc.subject | Machine learning | |
| dc.subject | Probabilistic | |
| dc.subject | Human factors | |
| dc.title | Probabilistic modeling with human factors in machine learning | |
| dc.type | Thesis | |
| dc.type.material | text | |
| thesis.degree.department | Computer Sciences | |
| thesis.degree.discipline | Computer Science | |
| thesis.degree.grantor | The University of Texas at Austin | |
| thesis.degree.level | Doctoral | |
| thesis.degree.name | Doctor of Philosophy |