Features

The new poll addresses the question of favorite format for datasets being mined. Of course data is intimately connected with meta-data, but past experience has shown that asking more than one question is reducing the response significantly.

Thus the question of metadata formats will be left to the next poll, and this poll asks:

What dataset format you use the most when data mining?
Choose one from

• Comma-separated (.csv) file
• Tab or Space separated (.txt) file
• Commercial data mining tool format (such as SAS, SPSS, etc.)
• Weka format (.ARFF)
• in a database (i.e. you keep datasets in a database)
• in a spreadsheet
• Other -- please comment.