Title | Discovering Process Models from Patient Notes |
---|
Authors | Banziger, R.B., Basukoski, A. and Chaussalet, T.J. |
---|
Type | Conference paper |
---|
Abstract | Process Mining typically requires event logs where each event is labelled with a process activity. That’s not always the case, as many process-aware information systems store process-related information in the form of text notes. An example are patient information systems (PIS), which store much information in the form of free-text patient notes. Labelling text-based events with their activity is not trivial, because of the amount of data involved, but also because the activity represented by a text note can be am-biguous. Depending on the requirements of a process analyst, we might need to label events with more or fewer unique activities: two similar events could represent the same activity (e.g. screen referral) or two different activities (e.g. screen adult ADHD referral and screen depression referral). We can therefore view activities as ontologies with an arbitrary number of entries. This paper proposes a method that produces an ontology for the activities of a process by analysing a text-based event log. We implemented an interactive tool that generates process models based on this ontology and the text-based event log. We demonstrate the proposed method’s usefulness by dis-covering a mental health referral process model from real-world data. |
---|
Keywords | Process Mining, Text Mining, Healthcare, Mental Healthcare. |
---|
Year | 2023 |
---|
Conference | 23rd International Conference on Computational Science (ICCS-23) |
---|
Publisher | Springer |
---|
Accepted author manuscript | |
---|
Publication dates |
---|
Published | 26 Jun 2023 |
---|
Journal | Lecture Notes in Computer Science |
---|
Journal citation | 10475, pp. 243-249 |
---|
ISSN | 1611-3349 |
---|
| 0302-9743 |
---|
Book title | Computational Science – ICCS 2023. ICCS 2023. Lecture Notes in Computer Science, vol 10475 |
---|
ISBN | 9783031360237 |
---|
| 9783031360244 |
---|
Digital Object Identifier (DOI) | https://doi.org/10.1007/978-3-031-36024-4_18 |
---|