Sentiment Analysis using Excel-Spreadsheet

TextMiner123
TextMiner123 New Altair Community Member
edited November 5 in Community Q&A
Hi Guys,

Hopefully you can give me some hints, as I dont know how to go on.
I have an Excel-Spreadsheet with Training Data containing two columns-One with the message, the other is the sentiment value (Positive or Negative).
For the training data, I used the process "Read Excel"--> " Data to Documents"--> "Process Documents"--> "Process Documents from Files"-->"W-J48".
I labeled the row "Sentiment value" on "Read Excel". But I always get the message "Input Example Set does not have a label attribute". I have no labels and directories in "Process Documents from Files". (as I have no single text files in directories). Therefore I build TextObjects on "Process Documents".
I have only a single Excel-Spreadsheet with thousands of messages including a column with the value "Positive" or "Negative" for each message. I want to apply machine learning. Any hints are welcome how to solve my problem.

Thanks and nice regards,
Brick

Answers

  • MariusHelf
    MariusHelf New Altair Community Member
    Hi Brick,

    do you know the video tutorial on the Rapid-I website? http://rapid-i.com/content/view/189/212/lang,en/ They offer a great possibility to get in touch with RapidMiner and DataMining.

    For your specific problem concerning the label: Use the Set Role operator to define the column with negative/positive in it as label.

    Best Regards,
    Marius