N-grams & Clusters

Under construction.

The N-gram and Cluster Frequency page lets you explore common word/tag sequences (n-grams) and search for clusters containing specific words or tags in your corpus.

What You Can Do

  • Generate frequency tables for n-grams (sequences of 2, 3, or 4 tags)
  • Search for clusters containing a specific word or tag
  • Filter n-grams by tag at each position
  • Download your results as an Excel file

Step 1: Choose Table Type

When you first visit the page, you’ll be asked:

  • N-grams: Find the most frequent sequences of tags (e.g., POS or DocuScope tags).
  • Clusters: Find all n-grams that contain a specific word or tag at a chosen position.
Important

What is an “n-gram”?
An n-gram is a sequence of n items (words or tags) that appear together in your text. For example, a 3-gram (trigram) could be “in the house” or “NOUN VERB NOUN”.

Step 2: Set Your Options

For N-grams

  • Span: Choose the length (2, 3, or 4).
  • Tagset: Choose between Parts-of-Speech or DocuScope tags.
  • Click N-grams Table to generate the table.

For Clusters

  • Search mode: Choose to search by a specific token (word) or tag.
  • Node word/tag: Enter the word or select the tag you want to anchor your search.
  • Search type (for tokens): Choose Fixed, Starts with, Ends with, or Contains.
  • Tagset: Choose Parts-of-Speech or DocuScope.
  • Span & position: Choose the n-gram length and the position of your anchor.
  • Click Clusters Table to generate the table.

Step 3: Filter and Explore

  • Once your table is generated, use the filters at the top to focus on specific tags at each position in the n-gram.
  • The table updates automatically as you filter.
Tip

Tip:
Filtering by tag position helps you find patterns, like all trigrams starting with a verb or ending with a specific rhetorical tag.

Step 4: Download Your Table

  • Toggle Download to Excel? in the sidebar to enable download.
  • Click Download to Excel to save your results for further analysis.

Generating a New Table

  • Use the Create a New Ngrams Table button in the sidebar to reset and start over with new settings.

Tips for New Users

Tip
  • Try generating both n-gram and cluster tables to see which is most useful for your research.
  • If you don’t see any data, make sure you have loaded and processed a target corpus.
  • Download your results often so you can experiment without losing your work.

If You Get Stuck

Important
  • Use the reset button on the Manage Corpus Data page if you need to start over.
  • If you see warnings, check that your corpus is loaded and processed, and that your search settings are valid.