Detecting order-disorder transitions in discourse: implications for schizophrenia.

Area of Science:

• Computational linguistics
• Network theory
• Psychiatric research

Background:

• Psychiatric and neurological conditions impact speech semantic organization.
• Schizophrenia discourse is often characterized by a lack of coherence.
• Evaluating discourse disturbances is crucial for diagnosis, treatment efficacy, and prognosis.

Purpose of the Study:

• To present novel computational tools for analyzing semantic coherence in patient discourse.
• To provide a foundation for automatic procedures in assessing communication patterns.
• To illustrate the application of these tools to speech transcriptions.

Main Methods:

• Review of information science and complex network approaches to semantic coherence.
• Development of a discourse representation for computing disorganization measures.
• Application of developed tools to speech data from patients and healthy individuals.

Main Results:

• Demonstration of new computational tools derived from network theory and information science.
• Illustration of how these tools can quantify semantic deviations in discourse.
• Empirical application to speech transcriptions, highlighting potential for clinical studies.

Conclusions:

• The presented framework offers a novel approach to measuring semantic disorganization in patient speech.
• These computational tools can aid empirical and clinical studies of communication patterns.
• The framework lays the groundwork for future automatic procedures in discourse analysis.