KeyGraph -Analysis procedure-

Analysis procedure

　By calculating the frequency of appearance of each of the elements that make up a certain piece of data, and the number and strength of the connections between them, KeyGraph categorizes and extracts those elements that play an important role. Furthermore, by visualizing the result as a network graph, it supports the discovery of the ideas concealed within the data.

Splitting into morphemes

When analyzing text, it is necessary to prepare the data format so that the morphemes that constitute the elements that make up the text can be detected.
KeyGraph includes functions to process normal text and split it into morphemes.

Survey results

・Text entered in forums and bogs, etc.
・POS data
・ Access logs
・ Weather data
etc.

Extract significant elements

Roof
ndicates an expression that is a potential opinion within the text, strongly connected to part of the words that form the foundation and to the foundation itself.

Pillar
In the main development of the text, indicates a connection between the base and roof.

Foundation
In the basic concept on which the text is based, indicates an expression that appears with high frequency.

Extracts the particularly important elements that make up the data, and calculates the number and strength of the connections between them.
KeyGraph analyzes the structure of the data based on the concepts of “Foundation,” “Roof,” and “Pillar” and so on.

Easily understood display

Displays a network diagram based on the “foundation,” “roof,” “pillars” and the keywords derived from the analysis. Morphemes are represented individually by nodes and the links between those nodes.