chriskirkhub/etea: This package enables the classification of unstructured textual data into a structured, segmented, temporal, document frequency matrix for use as input into predictive modelling systems such as neural networks or state-space models.

In order to analyse events that involve human behaviour or the interactions between human and organisation in diverse fields such as in special investigations, clinical treatment systems or capital market transactions, investigators benefit from examination of the content of clinical, case or transaction notes and records. Examination typically comprises the application of methodologies such as feature-extraction, classification and 'sentiment' valence analysis. By pre-processing unstructured text into components that also group transitions together, the potential for added value derived from adaptive algorithms processing and post-processing can be leveraged significantly to provide a multi-dimensional classification of textual content over state or time. The majority of cases can be described using classification categories grouped into Environment, Thought, Emotion and Action, (etea) from which the title of this package has been derived. Notably, these groupings also promote temporal analysis. This package enables the classification of unstructured textual data into a structured, segmented, temporal frequency matrix for use as input into predictive modelling systems such as neural networks or state- space models. This package is expected to be of interest to all those who seek to enhance knowledge discovery and to model using unstructured text as the data source. Typical cases include: behavioural analysis and segmentation studies of markets; scoring of texts for sequencing models or signature channels; special interest groups and investigations: Families with Complex Needs; Special Victims Units and Victim Support; Clinical or Treatment Systems.

install.packages("remotes")
remotes::install_github("chriskirkhub/etea")