Category:Text Analysis

Text Analysis can be done by selecting Create > Text Analysis. For an overview of text analysis in Q, see our case study. All options in the menu, except the word cloud, use R ().

The initial stage of most text analyses is to do some pre-processing of the text, which includes breaking the text up into separate words and doing some basic cleaning. These steps are performed by Create > Text Analysis > Setup Text Analysis. This creates an output in the Report which can then be used as an input to the other analyses from the Techniques section of the menu. The text processing is designed to be used with English-language text.

Word Cloud - Create a Word Cloud from an existing Text question. No setup is required, although you can make use of your processing by first using the Save Tidied Text option listed below.

Setup Text Analysis - Cleans text data for further analysis. Apply spell-checking, stemming, and manual replacement and removal of words.

Techniques

Create Term Document Matrix - Create a term document matrix for use in further analysis.

Predictive Tree - Create a predictive tree based on the words identified in the text.

Save Sentiment Scores - Save the sentiment scores for each case, based on a selected object created using Setup Text Analysis, or from a table containing raw text data.

Save Tidied Text - Save cleaned text data by case of a selected object created using Setup Text Analysis.

Search - Searches for a word or phrase and displays the cases containing the search term.

The Term Document Matrix is particularly useful as an input to other statistical algorithms, by writing your own R code.