자동 범주유창성검사 평가를 향하여: 단어 군집화를 활용한 그룹간 구별

Abstract

The CFT is a set of highly standardized assessments of cognitive function that are used in clinical neuropsychological testing throughout the world. The task is simple: Test subjects are asked to produce as many words from a given semantic category as possible during 60 seconds, where the most commonly used category is animals. The test mainly assesses verbal fluency and semantic memory, but performance is also affected by other aspects of executive function such as attention [1,2]. The standard way of scoring the test is to count the number of items generated. More in-depth scoring methods [3,4] build on a conceptual model of item generation that includes two components, or clusters and switches. Participants produce a cluster of items from the same subcategory. For example, they may produce a list of animal names which belong to canine such as wolf and fox. Once this is exhausted, they switch to another readily available cluster like those of pets as described in Figure 1.