Veronika Laippala1, Timo Viljanen2, Antti Airola3, Jenna Kanerva4, Sanna Salanterä5, Tapio Salakoski6, Filip Ginter7. 1. Department of French Studies, University of Turku, FI-20014, Finland. Electronic address: veronika.laippala@utu.fi. 2. Department of Information Technology, University of Turku, FI-20014, Finland. Electronic address: timo.viljanen@utu.fi. 3. Department of Information Technology, University of Turku, FI-20014, Finland. Electronic address: antti.airola@utu.fi. 4. Department of Information Technology, University of Turku, FI-20014, Finland. Electronic address: jenna.kanerva@utu.fi. 5. Department of Nursing Science, University of Turku, FI-20014, Finland; The Hospital District of Southwest Finland, PL 52, FI-20521 Turku, Finland. Electronic address: sanna.salantera@utu.fi. 6. Department of Information Technology, University of Turku, FI-20014, Finland. Electronic address: tapio.salakoski@utu.fi. 7. Department of Information Technology, University of Turku, FI-20014, Finland. Electronic address: filip.ginter@utu.fi.
Abstract
OBJECTIVES: In this paper, we study the development and domain-adaptation of statistical syntactic parsers for three different clinical domains in Finnish. METHODS AND MATERIALS: The materials include text from daily nursing notes written by nurses in an intensive care unit, physicians' notes from cardiology patients' health records, and daily nursing notes from cardiology patients' health records. The parsing is performed with the statistical parser of Bohnet (http://code.google.com/p/mate-tools/, accessed: 22 November 2013). RESULTS: A parser trained only on general language performs poorly in all clinical subdomains, the labelled attachment score (LAS) ranging from 59.4% to 71.4%, whereas domain data combined with general language gives better results, the LAS varying between 67.2% and 81.7%. However, even a small amount of clinical domain data quickly outperforms this and also clinical data from other domains is more beneficial (LAS 71.3-80.0%) than general language only. The best results (LAS 77.4-84.6%) are achieved by using as training data the combination of all the clinical treebanks. CONCLUSIONS: In order to develop a good syntactic parser for clinical language variants, a general language resource is not mandatory, while data from clinical fields is. However, in addition to the exact same clinical domain, also data from other clinical domains is useful.
OBJECTIVES: In this paper, we study the development and domain-adaptation of statistical syntactic parsers for three different clinical domains in Finnish. METHODS AND MATERIALS: The materials include text from daily nursing notes written by nurses in an intensive care unit, physicians' notes from cardiology patients' health records, and daily nursing notes from cardiology patients' health records. The parsing is performed with the statistical parser of Bohnet (http://code.google.com/p/mate-tools/, accessed: 22 November 2013). RESULTS: A parser trained only on general language performs poorly in all clinical subdomains, the labelled attachment score (LAS) ranging from 59.4% to 71.4%, whereas domain data combined with general language gives better results, the LAS varying between 67.2% and 81.7%. However, even a small amount of clinical domain data quickly outperforms this and also clinical data from other domains is more beneficial (LAS 71.3-80.0%) than general language only. The best results (LAS 77.4-84.6%) are achieved by using as training data the combination of all the clinical treebanks. CONCLUSIONS: In order to develop a good syntactic parser for clinical language variants, a general language resource is not mandatory, while data from clinical fields is. However, in addition to the exact same clinical domain, also data from other clinical domains is useful.