A Simple Topic Modeling Tool

I used a shortcut, rather than use the cumbersome (yeah, I said it) Mallet and the easy, but still not easy compared to the tools that should be out there, R language. My problem isn’t with Mallet and R themselves, it’s just that for the task at hand something easier needs to be out there.  In other words there should be something that can: upload a text file, set some parameters, and click GO…all without the command line, and all without coding. I’m uninterested in hearing about how the command line is easy (it is, but doesn’t matter), because using the command line has fallen out of the general vocabulary of mainstream computer usage, and digital humanists shouldn’t have to deal with it

 

I get the sense that we are on the brink of such tools being developed, along with a host of other tools that overlay a Graphical User Interface (GUI) on top of current mining, modeling, and analysis tools that would otherwise require coding, command lines, etc.  I tracked down one here.  Downloaded it, made sure I had Java, and ran it.  No Command Line, no Mallet, no R…just upload a text file, pick a few parameters, and click Learn Topics. I had a PDF of Davinci’s Notebooks, so I copied the whole thing and pasted it into textedit to save it all as a .txt file, which seems to be the kind of file this tool requires.  Then I selected the .txt file within the little program and viola.  Take a peek at the set up of the program:

 

 

So, the topics it came up with were:

 

1,time move sea earth equal line small distance animals world

2,part body make sun thing side ing dark head turned

3,man parts arts hand height colours works movements drawing means

4,movement weight power force wings bird greater moves infinite birds

5,nature leonardo men painting good day left period florence set

6,great mountains colour back trees clouds high river waters made

7,eye light object painter give point number found shadows times

8,water air wind surface centre elements fire natural leaves element

9,things life bodies objects images form human reason find front

10,made work order milan study sforza forms years house weights

 

It also exports all of the statistical data in csv format so that this data can be easily used with data visualization tools like D3 or simple ones like excel (shudder).

 

Now, as to what I make of those topics? Damn, is there a program for THAT?  🙂

Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s