This output format is designed to create XY-scatterplots in comma-separated-value (CSV) format for use in standard spreadsheet applications like Microsoft Excel or OpenOffice Calc, which can import CSV and create scatterplots from the data.
This option is available for Term-chunk and Chunk-term searches. It is intended to help visualize the distribution of terms across the corpus in terms of their cosine correlations with the chunks.
This is more than simple term frequencies. The cosine correlations between terms and chunks convery real information about how important that chunk is to the expression or definition of that term in this corpus.
Terms can be present in locations but not be important there or their use there may be insignificant relative to other passages. The LSA calculation actually provides a measure of the locatedness of terms within the corpus.