Computational linguistic analysis applied to a semantic fluency task to measure derailment and tangentiality in schizophrenia

Abstract Human- or expert-generated records that describe the behavior of engineered systems over a period of time can be useful for statistical learning techniques like pattern detection or output prediction. However, such data often assumes familiarity of a reader with the relationships between entities within the system — that is, knowledge of the system’s structure. This required, but unrecorded “tacit” knowledge makes it difficult to reliably learn patterns of system behavior using statistical modeling techniques on these written records. Part of this difficulty stems from a lack of good models for how engineers generate written records of a system, given their expertise, since they often create such records under time pressure using shorthand notation or internal jargon. In this paper, we model the process of maintenance work order creation as a modified semantic fluency task, to build a probabilistic generative model that can uncover underlying relationships between entities referenced within a complex system. Compared to more traditional similarity-metric-based methods for structure recovery, we directly model a possible cognitive process by which technicians may record work-orders. Mathematically, we represent this as a censored local random walk over a latent network structure representing tacit engineering knowledge. This allows us to recover implied engineering knowledge about system structure by processing written records. Additionally, we show that our model leads to improved generative capabilities for synthesizing plausible data.

Download Full-text

The effect of age and sex on clustering and switching during speeded verbal fluency tasks

Journal of the International Neuropsychological Society ◽

10.1017/s1355617709090237 ◽

2009 ◽

Vol 15 (2) ◽

pp. 196-204 ◽

Cited By ~ 62

Author(s):

SHAWNDA LANTING ◽

NICOLE HAUGRUD ◽

MARGARET CROSSLEY

Keyword(s):

Older Adults ◽

Sex Differences ◽

Verbal Fluency ◽

Verbal Ability ◽

Past Research ◽

Strategy Use ◽

Semantic Fluency ◽

Age Related ◽

Fluency Task ◽

Age And Sex

AbstractPast research has been inconsistent with regard to the effects of normal aging and sex on strategy use during verbal fluency performance. In the present study, both Troyer et al.’s (1997) and Abwender et al.’s (2001) scoring methods were used to measure switching and clustering strategies in 60 young and 72 older adults, equated on verbal ability. Young adults produced more words overall and switched more often during both phonemic and semantic fluency tasks, but performed similarly to older adults on measures of clustering. Although there were no sex differences in total words produced on either fluency task, males produced larger clusters on both tasks, and females switched more frequently than males on the semantic but not on the phonemic fluency task. Although clustering strategies appear to be relatively age-insensitive, age-related changes in switching strategies resulted in fewer overall words produced by older adults. This study provides evidence of age and sex differences in strategy use during verbal fluency tests, and illustrates the utility of combining Troyer’s and Abwender’s scoring procedures with in-depth categorization of clustering to understand interactions between age and sex during semantic fluency tasks. (JINS, 2009, 15, 196–204.)

Download Full-text

Influence of age on the dynamics of fMRI activations during a semantic fluency task

Journal of Neuroradiology ◽

10.1016/j.neurad.2011.05.005 ◽

2012 ◽

Vol 39 (3) ◽

pp. 158-166 ◽

Cited By ~ 13

Author(s):

C. Destrieux ◽

C. Hommet ◽

F. Domengie ◽

J.-M. Boissy ◽

G. De Marco ◽

...

Keyword(s):

Semantic Fluency ◽

Fluency Task

Download Full-text

Risk Factors for Selective Cognitive Decline in Dialyzed Patients with End-Stage Renal Disease: Evidence from Verbal Fluency Analysis

Journal of the International Neuropsychological Society ◽

10.1017/s1355617711001445 ◽

2011 ◽

Vol 18 (1) ◽

pp. 162-167 ◽

Cited By ~ 20

Author(s):

Michał Harciarek ◽

John B. Williamson ◽

Bogdan Biedunkiewicz ◽

Monika Lichodziejewska-Niemierko ◽

Alicja Dębska-Ślizień ◽

...

Keyword(s):

Renal Disease ◽

End Stage Renal Disease ◽

Semantic Fluency ◽

Temporal Lobes ◽

Phonemic Fluency ◽

Stage Renal Disease ◽

End Stage ◽

Cognitive Problems ◽

Fluency Task ◽

Direct Confirmation

AbstractAlthough dialyzed patients often have cognitive problems, little is known about the nature of these deficits. We hypothesized that, in contrast to semantic fluency relying mainly on temporal lobes, phonemic fluency, preferentially depending on functions of frontal-subcortical systems, would be particularly sensitive to the constellation of physiological pathological processes associated with end-stage renal disease and dialysis. Therefore, we longitudinally compared phonemic and semantic fluency performance between 49 dialyzed patients and 30 controls. Overall, patients performed below controls only on the phonemic fluency task. Furthermore, their performance on this task declined over time, whereas there was no change in semantic fluency. Moreover, this decline was related to the presence of hypertension and higher blood urea nitrogen. We suggest that these findings may be due to a combination of vascular and topic effects that impact more on fronto-subcortical than temporal lobe networks, but this speculation requires direct confirmation. (JINS, 2012,18, 162–167)

Download Full-text

A computational linguistic measure of clustering behavior on semantic verbal fluency task predicts risk of future dementia in the Nun Study

Cortex ◽

10.1016/j.cortex.2013.05.009 ◽

2014 ◽

Vol 55 ◽

pp. 97-106 ◽

Cited By ~ 20

Author(s):

Serguei V.S. Pakhomov ◽

Laura S. Hemmy

Keyword(s):

Computational Linguistic ◽

Verbal Fluency ◽

Verbal Fluency Task ◽

Semantic Verbal Fluency ◽

Nun Study ◽

Clustering Behavior ◽

Fluency Task

Download Full-text

Comparing the semantic networks of children with cochlear implants and children with typical hearing: Extending Kenett et al. (2013) by considering Hearing Age.

10.31234/osf.io/56khg ◽

2021 ◽

Author(s):

Michaela Socher ◽

ulrika löfkvist ◽

Malin Wass

Keyword(s):

Cochlear Implants ◽

Semantic Network ◽

Semantic Networks ◽

Spoken Language ◽

Chronological Age ◽

Perceptual Quality ◽

Semantic Fluency ◽

Verbal Fluency Task ◽

Significant Difference ◽

Fluency Task

Purpose: Kenett et al. (2013) report that the sematic network of children with CI is less structured compared to the sematic network of children with TH. This study aims to evaluate if such differences are only evident if children with CI are compared to children with TH matched on chronological age, or also if they are compared to children with TH matched on hearing age. Method: The performance of a group of children with CI on a verbal fluency task was compared to the performance of a group of chronological-age matched children with TH. Subsequently, computational network analysis was used to compare the semantic network structure of the groups. The same procedure was applied to compare a group of children with CI to a group of hearing-age matched children with TH. Results: Children with CI performed significantly more poorly than children with TH matched on chronological age on a semantic fluency task and exhibited a significantly less structured semantic network. No significant difference in performance on a semantic fluency task was found between children with CI and children with TH matched on hearing-age. However, the structure of the semantic network differed significantly for the hearing age matched groups. Conclusions: Although the groups perform on the same level on a sematic fluency task, the semantic network for spoken language of children with CI is less structured compared to children with TH matched on hearing age. Reasons for this might be differences in the (perceptual) quality and the quantity of spoken language input.

Download Full-text

Viability in Multiplex Lexical Networks and Machine Learning Characterizes Human Creativity

Big Data and Cognitive Computing ◽

10.3390/bdcc3030045 ◽

2019 ◽

Vol 3 (3) ◽

pp. 45 ◽

Cited By ~ 8

Author(s):

Massimo Stella ◽

Yoed N. Kenett

Keyword(s):

Machine Learning ◽

Individual Differences ◽

Language Learning ◽

Conceptual Knowledge ◽

Area Under The Curve ◽

Mental Lexicon ◽

Network Models ◽

Semantic Fluency ◽

Multiplex Network ◽

Fluency Task

Previous studies have shown how individual differences in creativity relate to differences in the structure of semantic memory. However, the latter is only one aspect of the whole mental lexicon, a repository of conceptual knowledge that is considered to simultaneously include multiple types of conceptual similarities. In the current study, we apply a multiplex network approach to compute a representation of the mental lexicon combining semantics and phonology and examine how it relates to individual differences in creativity. This multiplex combination of 150,000 phonological and semantic associations identifies a core of words in the mental lexicon known as viable cluster, a kernel containing simpler to parse, more general, concrete words acquired early during language learning. We focus on low (N = 47) and high (N = 47) creative individuals’ performance in generating animal names during a semantic fluency task. We model this performance as the outcome of a mental navigation on the multiplex lexical network, going within, outside, and in-between the viable cluster. We find that low and high creative individuals differ substantially in their access to the viable cluster during the semantic fluency task. Higher creative individuals tend to access the viable cluster less frequently, with a lower uncertainty/entropy, reaching out to more peripheral words and covering longer multiplex network distances between concepts in comparison to lower creative individuals. We use these differences for constructing a machine learning classifier of creativity levels, which leads to an accuracy of 65 . 0 ± 0 . 9 % and an area under the curve of 68 . 0 ± 0 . 8 % , which are both higher than the random expectation of 50%. These results highlight the potential relevance of combining psycholinguistic measures with multiplex network models of the mental lexicon for modelling mental navigation and, consequently, classifying people automatically according to their creativity levels.

Download Full-text