It's Difficult to be Neutral -- Human and LLM-based Sentiment Annotation of Patient Comments
This work addresses the resource-intensive problem of sentiment annotation in healthcare for improving patient feedback analysis, but it is incremental as it evaluates existing LLMs on new data.
The study tackled the challenge of sentiment annotation for patient comments by comparing human annotators with large language models (LLMs) on a Norwegian dataset, finding that LLMs performed well above baseline in zero-shot binary sentiment tasks but still fell short of human performance on the full dataset.
Sentiment analysis is an important tool for aggregating patient voices, in order to provide targeted improvements in healthcare services. A prerequisite for this is the availability of in-domain data annotated for sentiment. This article documents an effort to add sentiment annotations to free-text comments in patient surveys collected by the Norwegian Institute of Public Health (NIPH). However, annotation can be a time-consuming and resource-intensive process, particularly when it requires domain expertise. We therefore also evaluate a possible alternative to human annotation, using large language models (LLMs) as annotators. We perform an extensive evaluation of the approach for two openly available pretrained LLMs for Norwegian, experimenting with different configurations of prompts and in-context learning, comparing their performance to human annotators. We find that even for zero-shot runs, models perform well above the baseline for binary sentiment, but still cannot compete with human annotators on the full dataset.