Hi,

You would need to convert your documents to tf-idf vectors, remove all
stopwords and run rowSimilarity on that with cosine as similarity
measure. That should you give you reasonable results.

--sebastian

Am 26.11.2010 06:27, schrieb Divya:
> Hi,
> 
>  
> 
> I need to know what is the usage of rowSimilarity CL.
> 
> I know we use I compute the pairwise row similarity.
> 
> I want to know more about it.
> 
> Where we can use it .
> 
> Can we use it compute similarity between two documents contents.
> 
>  
> 
> Regards,
> 
> Divya 
> 
> 

Reply via email to