XLSTAT: Feature Extraction in Excel

 

Background

Feature extraction is used to reduce the number of resources required to describe a large set of textual data. It is a general term for methods of constructing combinations of the variables to get around these problems while still describing the data with sufficient accuracy.

The  extracted features are commonly used in methods of document classification, where the frequency of occurrence of each word in a document is used as a feature for training a classifier.

 

Dataset for Example

The attached data are from the  Internet Movie Database (IMDB) which consists of 4000 movie reviews written in English.

 

Setting Up

• Open XLSTAT.

• Click on the Marketing or Sensory menu/Text mining and select the Feature extraction feature.

• The Feature Exxtraction dialog box will appear. 

In the General tab:

• Select the column "Review" in the Worksheet field

• Select the column "Id" in the Document labels  field.

In the Options / Preprocessing  sub-menu:

• Exclude a list of English stop words.

• Remove punctuation marks and numbers.

• Activate the Stemming option.

In the Intermediate form sub-menu, apply filtering:

• Activate the Remove sparse terms  option.

• Enter 10 as the  Minimum frequency.

  

• Click on the OK button to launch the computations.

• The results are displayed on a new Excel sheet.

 

Interpreting the Results

The document term-matrix is displayed.

Terms present in less than 5% of all reviews or appearing fewer than 10 times were excluded from the document-term matrix.

A word cloud showing the most frequent terms across all reviews follows the matrix.