Preprocess Text Data

Preprocess and clean up text data for analysis

Since R2023a

expand all in page

Description

The Preprocess Text Data Live Editor task helps prepare text data for analysis.

You can use the task to control these processing steps:

  • HTML clean up

  • Tokenization

  • Adding token details

  • Word normalization

  • Changing and removing words

The Preprocess Text Data Live Editor task generates code that performs the selected preprocessing steps, which you can use to create a preprocessing function for your workflows.

Preprocess Text Data Task in Live Editor

Open the Task

To add the Preprocess Text Data task to a live script in the MATLAB® Editor:

  • On the Live Editor tab, select Task > Preprocess Text Data.

  • In a code block in the live script, type a relevant keyword, such as preprocess, clean, or text. Select Preprocess Text Data from the suggested command completions.

Examples

expand all

This example shows how to create a function which cleans and preprocesses text data for analysis using the Preprocess Text Data Live Editor task.

First, load the factory reports data. The data contains textual descriptions of factory failure events.

tbl = readtable("factoryReports.csv")

A table with variables "Description", "Category", and "Urgency". The "Description" variable contains textual descriptions such as "Items are occasionally getting stuck in the scanner spools". The "Category" variable contains categorical labels such as "Mechanical Failure", and the "Urgency" variable contains categorical labels such as "Medium".

Open the Preprocess Text Data Live Editor task. To open the task, begin typing the task name and select Preprocess Text Data from the suggested command completions. Alternatively, on the Live Editor tab, select Task > Preprocess Text Data.

Drop down list showing suggested command completions. The only suggestion in the list is for the Preprocess Text Data task, and is selected.

Preprocess the text using these options:

  1. Select tbl as the input data and select the table variable Description.

  2. Tokenize the text using automatic language detection.

  3. To improve lemmatization, add part-of-speech tags to the token details.

  4. Normalize the words using lemmatization.

  5. Remove words with fewer than 3 characters or more than 14 characters.

  6. Remove stop words.

  7. Erase punctuation.

  8. Display the preprocessed text in a word cloud.

Preprocess Text Data task with fields corresponding to preprocessing options highlighed with numbered red rectangles. The image highlights these options in order: "Data", "Language", "Add part-of-speech tags", "Normalize words", "Minimum word length", "Maximum word length", "Remove stop words", "Erase punctuation", and "Show word cloud".

A word cloud showing words in different font sizes. Larger font sizes indicate more frequent words in the data. The word cloud highlights words like "assembler" and "mixer". Words like "the" and "in" do not appear in the word cloud.

The Preprocess Text Data Live Editor task generates code in your live script. The generated code reflects the options that you select and includes code to generate the display. To see the generated code, click Show code at the bottom of the task parameter area. The task expands to display the generated code.

MATLAB code generated by Preprocess Text Data task

By default, the generated code uses preprocessedText as the name of the output variable returned to the MATLAB workspace. To specify a different output variable name, enter a new name in the summary line at the top of the task.

First option of the Preprocess Text Data task

To reuse the same steps in your code, create a function that takes as input the text data and outputs the preprocessed text data. You can include the function at the end of a script or as a separate file. The preprocessTextData function listed at the end of the example, uses the code generated by the Preprocess Text Data Live Editor task.

To use the function, specify the table as input to the preprocessTextData function.

documents = preprocessTextData(tbl);

Preprocessing Function

The preprocessTextData function uses the code generated by the Preprocess Text Data Live Editor task. The function takes as input the table tbl and returns the preprocessed text preprocessedText. The function performs these steps:

  1. Extract the text data from the Description variable of the input table.

  2. Tokenize the text using tokenizedDocument.

  3. Add part-of-speech details using addPartOfSpeechDetails.

  4. Lemmatize the words using normalizeWords.

  5. Remove words with 2 or fewer characters using removeShortWords.

  6. Remove words with 15 or more characters using removeLongWords.

  7. Remove stop words (such as "and", "of", and "the") using removeStopWords.

  8. Erase punctuation using erasePunctuation.

function preprocessedText = preprocessTextData(tbl)

%% Preprocess Text
preprocessedText = tbl.Description;

% Tokenize
preprocessedText = tokenizedDocument(preprocessedText);

% Add token details
preprocessedText = addPartOfSpeechDetails(preprocessedText);

% Change and remove words
preprocessedText = normalizeWords(preprocessedText,Style="lemma");
preprocessedText = removeShortWords(preprocessedText,2);
preprocessedText = removeLongWords(preprocessedText,15);
preprocessedText = removeStopWords(preprocessedText,IgnoreCase=false);
preprocessedText = erasePunctuation(preprocessedText);

end

For an example showing a more detailed workflow, see Preprocess Text Data in Live Editor. For next steps in text analytics, you can try creating a classification model or analyze the data using topic models. For examples, see Create Simple Text Model for Classification and Analyze Text Data Using Topic Models.

Parameters

expand all

Select Data

Text to preprocess, specified as a MATLAB workspace variable. The variable must be a table, string array, or character vector to appear in the list.

If you select a table, then specify the table variable containing the text data in the second drop-down box that appears.

Clean Up HTML

Extract text data from HTML tags.

The generated code uses extractHTMLText.

Remove HTML tags.

The generated code uses eraseTags.

Convert HTML and XML entities into characters. For example convert "&" to "&".

The generated code uses decodeHTMLEntities.

Tokenize

Text language, specified as one of these options:

Automatic

Automatic language detection

English

English language

German

German language

Japanese

Japanese language

Korean

Korean language

The generated code uses tokenizedDocument.

Text splitting mode, specified as one of these options:

None

Do not split input.

Sentences

Split input into sentences. This option supports scalar input only.

The generated code uses splitSentences.

Paragraphs

Split input into paragraphs. This option supports scalar input only.

The generated code uses splitParagraphs.

Add Token Details

Option to add sentence numbers to tokens.

The generated code uses addSentenceDetails.

Option to add part-of-speech tags to tokens.

The generated code uses addPartOfSpeechDetails.

Option to detect named entities in tokens.

The generated code uses addEntityDetails.

Option to parse dependencies in tokens. This option requires Text Analytics Toolbox™ Model for Udify data support package.

The generated code uses addDependencyDetails.

Token Edit and Removal

Word normalization, specified as one of these options:

None

Do not normalize words.

Lemma

Normalize words using lemmatization. This option outputs text in lowercase.

Stem

Normalize words using stemming.

The generated code uses normalizeWords.

Case normalization, specified as one of these options:

None

Do not normalize case.

Note

The Lemma option of Word normalization converts text to lowercase.

Lowercase

Convert text to lowercase.

The generated code uses lower.

Uppercase

Convert text to uppercase.

The generated code uses upper.

Minimum word length, specified as of these options:

  • off — Do not remove short words

  • positive integer — remove words with fewer than the specified number of characters

The generated code uses removeShortWords.

Maximum word length, specified as of these options:

  • off — Do not remove long words

  • positive integer — remove words with more than the specified number of characters

The generated code uses removeLongWords.

Option to remove stop words.

The generated code uses removeStopWords.

Option to erase punctuation.

The generated code uses erasePunctuation.

Source and target words for replacement, specified as pairs of source and target strings. To specify multiword phrases (n-grams), use whitespace separated words.

The generated code uses replaceWords and replaceNgrams.

Words to remove, specified as strings. To specify multiword phrases (n-grams), use whitespace separated words.

The generated code uses removeWords and removeNgrams.

Option to remove empty documents.

The generated code uses removeEmptyDocuments.

Option to ignore case in word change and removal options.

Display Results

Option to show tokenized text.

Option to show token details.

The generated code uses tokenDetails.

Option to show word cloud.

The generated code uses wordcloud.

Tips

  • By default, the Preprocess Text Data task does not automatically run when you modify the task parameters. To have the task run automatically after any change, select the Autorun checkbox at the top-right of the task. If your data set is large, do not enable this option.

Version History

Introduced in R2023a

See Also

| | | | | |

Topics