"Word-Pairs and Word-Triplets from Latent Semantic Analysis"

MacGregor

I have a question, you can find pairs and triples words from an Excel? In the model I am doing, I have a excel where documents are represented by columns and the rows for keywords with their frequency or co-ocurrence in each document. Then:

1 - Read the Excel (rows: words - columns: documents)
2 - the out in -> W-LatentSemanticAnalysis
3 - the exa in -> AglomerativeClustering output
4 - the exa in to -> Scale by Weights

So far so good, the graphics are very cool but how can i form pairs and triples of words?.. Someone has done something similar?. Appreciate any help ..

Thanks in advance.

ahootanha

I also want to use a loose clustering for meaning. Please explain the process?
Thanks

---

@MacGregor wrote:
I have a question, you can find pairs and triples words from an Excel? In the model I am doing, I have a excel where documents are represented by columns and the rows for keywords with their frequency or co-ocurrence in each document. Then:

1 - Read the Excel (rows: words - columns: documents)
2 - the out in -> W-LatentSemanticAnalysis
3 - the exa in -> AglomerativeClustering output
4 - the exa in to -> Scale by Weights

So far so good, the graphics are very cool but how can i form pairs and triples of words?.. Someone has done something similar?. Appreciate any help ..

Thanks in advance.

---

 

sgenzer

hello @ahootanha this is a VERY old thread from 2012. I would recommend starting a new thread...with a sample XML process and dataset.

 

Scott