RapidAnalytics and Tag Cloud

[mmarag]

Hello to all,

i have built a process that Reads RSS feeds, sends them to GetPages, open the link and then feeds it to a process documents from examples, where extract content, tokenize and filter lenght are applied.

I ran this from RapidAnalytics and wanted to have a tag cloud of the exctracted words. I can only achieve to obtain a table with all words and their term and document frequency (when i select PLOT parameter in the export as service). When i select tag cloud nothing happens. I do not know what are these parameters it asks for and how optional they are.

Any feedback?

Manolis

[Simon Fischer]

Hi,

you should at least set "attribute_term" attribute to something reasonable. Specify the attribute name which contains the term. "attribute_size" specifies the size of the words. Frequencies are not useful here, because they will be close to 0 and you don't want 0px sized words :-) Scale appropriately in RapidMiner.

Best,
Simon

[mmarag]

Dear Sir,

thank you very much, i managed to make it work.

Regards

MM

[mmarag]

Dear Sir,

i had an example set with 2 columns, word and total. Word contain terms and total contains integers (between 10 and 50).
when i try to export the process as service, i select TAG_CLOUD, attribute_tem=word and attribute_size=total.

it runs smoothly but i do not know the explanation of attribute_href,attribute_hicolor, and href_prefix.

Kind regards

Manolis

[Simon Fischer]

Hi,

If you use href, you will get links that you can click on. hicolor is the color (in hex) when highlighting. If your href attributes are not absolute, you can use the href_prefix to add before all links.

Best,
Simon