how find Cluster center in text

m_keshavarz_comm_keshavarz_com Member Posts: 28 Contributor I
edited December 2018 in Help

Hello to all
Sorry
I am a beginner at RapidMiner
I need to work with RapidMiner for the university project
I have a thousand reviews
I want to cluster with more important words by weighting
I did this process
But
Only the words weighed clustered
I want to make sentences based on these words clustered
But there is an error
Can anyone tell me what to do?
And
I want to find the center of the cluster
What is the word cluster center or sentence?
Does anyone know how?

Thanks a lot
Please help me

bb.JPGbb1.JPG

Answers

  • atamulewiczatamulewicz Member Posts: 21 Contributor II

    Hi @m_keshavarz_com,

     

    Can you post your XML? This way we can assist you. 

     

    Thanks, 

    Allie 

  • Thomas_OttThomas_Ott RapidMiner Certified Analyst, RapidMiner Certified Expert, Member Posts: 1,761 Unicorn

    @m_keshavarz_com your problem is that you're passing non-numerical data into the 'Determine Influence Factors' subprocess. Either Dummy-code them or remove them. 

  • m_keshavarz_comm_keshavarz_com Member Posts: 28 Contributor I

    Hello
    Dear Professor
    Mr. @Thomas_Ott
    Sorry i did not understand
    what must I do?
    Can you guide me more accurately?
    Thank you
    I really need your help.
    With respect

     

  • Thomas_OttThomas_Ott RapidMiner Certified Analyst, RapidMiner Certified Expert, Member Posts: 1,761 Unicorn

    @m_keshavarz_com you might to put a Nominal to Numerical operator before the data gets to the Influence Factors subprocess. This way all the data is numerical and the subprocess should be able to run. 

  • m_keshavarz_comm_keshavarz_com Member Posts: 28 Contributor I

    Hello
    Thank you
    You are right
    I did
    But clustering only the words that are more importantly extracted. Not the sentences containing the words
    I want clustered sentences based on these important words.
    But do not know how
    Thanks a lot

Sign In or Register to comment.