LogosLink User's Manual · LogosLink version 2.0.0

Lexical/Semantic Collocation Analytics (Argumentation Model)

Lexical/Semantic Collocation analytics shows what concepts appear close to a selected word in the argumentation model text content.

This is useful to find out major word/concept associations.

Parameters

Focus word. This is the word that you want to find across the argumentation model, and for which nearby concepts are to be determined.
Max distance. This is the maximum number of words before of after an occurrence of the focus word that will be considered to determine collocated concepts. Typical values are around 5.
Speaker. If you select a speaker, only the locutions of this speaker will be taken into account.
Agent. If the argumentation model has an embedded context and you select an agent, only the locutions of speakers linked to this agent will be taken into account.
Case sensitive. If you check this box, words that differ only in case (such as "word" and "Word") are treated as different words.

Results

Results are given as some overall data and a chart.

Overall data

Focus word occurrence count. This is the number of times that the focus word occurs in the text.
Collocation count. This is the number of concept occurrences that were found within the maximum distance of focus word occurrences.

Chart

The chart shows the collocated concepts along the vertical axis. The horizontal axis shows the distance of each collocated concept to the focus word. Blue bars show distance ranges, from minimum to maximum distance. Red circles show average distances, and their size is proportional to the number of occurrences of the collocated concept.

You can select how collocated concepts are sorted from the drop down menu in the toolbar.

Details

Collocated concepts are found in the text via denotations to the embedded ontology. The analytics looks for denotations within the maximum distance from each focus word occurrence in each locution, and presents the associated concepts as results.