Problems with visualizing a co-occurrence matrix in Kibana

[Matthew_1979 (Matthew)]

Hello everyone,

I am trying to visualize a co-occurrence matrix. This matrix consists of two lists with terms (list1: n_columns & list2: n_rows). The data itself shows the article identifiers in which the terms co-occur. So 'potassium' and 'copper' co-occur in the articles '31593825', '31579618' etc.

co-occurrence matrix:

alternative input -> single list (binary):

I tried to import and modify the data in Kibana in different ways, but I can't seem to find the right solution for this problem. The data in its current format looks un-visualizable to me and probably needs to be changed or saved in Elasticsearch in a specific way.

I want to create visualizations like a 'Wordcloud', 'Sunburst' or 'Co-occurrence network model'. Is there a solution to save the data in ES in such a way that I will be able to create all these visualizations? If not, only the 'Wordcloud'?

Thanks

[Nathan_Reese]

Have you tried storing your material values in a single field? For example, each document would have a field material with values like copper or zinc or so on. Fields can contain arrays so you can store multiple values in a single field.

By storing values in a single field, you can use the existing tag cloud visualization and vega to create sunburst visualization.

Then you could try using graph for co-occurrence network model.