Symptom Checklist-90-Revised Rating Scale: A Data Mining Approach
Evgenia Gkintoni1, Magdalena Nikiel2, Sampson Fytros2 Constantinos Halkiopoulos2,3, Gerasimos Antzoulatos2,3
[Abstract]
In this paper were applied Machine Learning and Data Mining methods to capture the psychological state group of students 18-26 years. For recording, tracing and evaluation of the psychological condition, was used the standardized scale Symptom Checklist-90 (SCL-90), which examines a wide range of psychological problems and symptoms of psychopathology.
The methodology adopted, in first phase consists of electronic questionnaires, which were created and posted through the website https://www.cicos.gr. Subsequently data were collected and preprocessed from the questionnaires and then introduced into the R (Programming Language and Machine Learning Platform) for analysis and extraction of useful knowledge. More specifically, through using classification algorithms (ID3, C4.5) there was a production of prospectively decision trees. Decision trees are a powerful way in order to represent and facilitate statements analysis (psychological) principally, comprising successive decisions and variable results in a designated period.
Furthermore, clustering technique (K-Means algorithm), was applied, which is a well-known knowledge discovery process of extracting previously unknown knowledge, actionable information from very large scientific and commercial databases. The k–means is a very popular algorithm and one of the best for implementing the clustering process. Also, the parameters of the algorithm were set, depending on the application cases, and also the results were correlated with the birth-place and the place of present residence, educational background of both the respondents and first-degree relatives, professional occupation of parents and other parameters, in order to evaluate and assess the significance of exported rules / conclusions. In addition, the respondents were classified into clusters based on 9 clinical signs (subscales) of the scale SCL-90.
The results indicate among others, that the use of Data Mining methods is an important tool to export and receive the conclusions and decisions especially in the field of psychological assessment and in neuroscience.
1Department of Psychology, University of Crete, Greece
2Department of Business Administration, Technological Educational Institute of Western Greece, Greece
3Department of Mathematics, University of Patras Artificial Intelligence Research Center, University of Patras , Greece
[Poster]
[Conference Program ECSP-2015]
© The Author(s) 2015. Published by ECSP 2015 – 2nd European Congress for Social Psychiatry Proceedings [1 July — 3 July 2015, UniMail, Geneva, Switzerland] on behalf of Cicos Research Team and the Departments of Business Administration (Technological and Educational Institute of Western Greek), Psychology (University of Crete) and Mathematics (University of Patras).
This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited.