Orange Forum • View topic - [HELP] Text Mining (bag of words)

[HELP] Text Mining (bag of words)

A place to ask questions about methods in Orange and how they are used and other general support.

[HELP] Text Mining (bag of words)

Postby Kamekjay » Fri Feb 15, 2013 17:24

Hi everyone, i'm new to Orange and i have to understand some things about the widget "bag of words" in the text mining module.
I have to visualize the results of the bag of words widget...i have a txt input correctly processed by preprocess widget and bag of words widget. the txt are transcription of some speeches and i have to find which words are most frequent and then visualize a sort of distance between them in the document.
First of all i have to extract the frequency about the words and eventually not considering common words like articles, etc...is it possible? please help me.

Re: [HELP] Text Mining (bag of words)

Postby Kamekjay » Mon Feb 18, 2013 14:20

please help me...i nedd to know how to visualize the results of the bag of words widget...i'm not understanding how it works.

Re: [HELP] Text Mining (bag of words)

Postby Ales » Mon Feb 18, 2013 18:23

Kamekjay wrote:... eventually not considering common words like articles, etc...is it possible
I think this is already handled (at least it should be) by the Text/Preproces widget.

Re: [HELP] Text Mining (bag of words)

Postby Kamekjay » Mon Feb 18, 2013 18:32

yes for that option i've already found the solution....now the problem is to visualize the distance of every text based on frequency of words...i think that mds is the solution but i have some trouble to understand the output

Re: [HELP] Text Mining (bag of words)

Postby Kamekjay » Thu Feb 21, 2013 17:29

Ales can i ask you an hint about visualization of data? i have a text correctly preprocessed with preprocessor and bag of words, and i would like to visualize all the words in the text with their frequency in the documents collection. can you suggest how can i do that? it would be nice to visualize them on histogram...

Re: [HELP] Text Mining (bag of words)

Postby roujri0005 » Fri Oct 11, 2013 11:26

Kamekjay wrote:Hi everyone, i'm new to Orange and i have to understand some things about the widget "bag of words" in the text mining module.
I have to visualize the results of the bag of words widget...i have a txt input correctly processed by preprocess widget and bag of words widget. the txt are transcription of some speeches and i have to find which words are most frequent and then visualize a sort of distance between them in the document.
First of all i have to extract the frequency about the words and eventually not considering common words like articles, etc...is it possible? please help me.

@Kamekjay I am trying the bag of words model basically a text file containing text information, however when I load the file I get an error in classifier and text learner. Did you link the data with bag of words model in you model. How your text file format is structured. Thanks.


Return to Questions & Support