[SOLVED] Text Mining

Rhmanig

Hi Everyone  Smiley

I am relatively new to RapidMiner and would appreciate if you could help me with this. I have a large data set (search log data) containing information as below:

userID, Query, ItemRank, ClickURL, QueryTime
(There are multiple records with same UerID)

I would like to use RapidMiner to extract useful information, (i.e, find individuals with specific skills. hard is it not!)

So far I have cleansed the date, and now would like to split the data based on UserID.

Then put data (queries) related to each UserID in to a document so I can analysis individual user's query. But I am not sure how?

I have retrieved the data from csv, then used DataToDocument to create a document from each record (rows), now would like to combine document with same UserID into one, but it seems not pussible? Huh

Can you please suggest some way to achieve this.

Regards

sgenzer

Hello.  Hmmm...that's not the way I would do that.  I would import the csv and then use the "Loop Values" operator on the example set where the attribute would be your UserID.  Inside the loop, use the macro "loop_value" to then filter the example set to just have that UserID.  Then do what you like.

Scott

MartinLiebig

Can't you simply use the aggregation first using concat and group by userID? then you would need to replace the added | with a \n or something and you are done.

If you provide example data, i could build an example process for you

Rhmanig

Thank you guys.