Batch registration of multiple words in the user dictionary for speech recognition

If you have multiple words that you want to register in the user dictionary for speech recognition, you can import them into a CSV file for efficient batch registration.


The maximum number of words to be registered in the user dictionary for speech recognition is 100.

Configuration of CSV file to be created

The CSV file to be used for importing should have the following structure.


Word 1, Reading 1, Part of Speech 1
Word 2, Reading 2, Part of Speech 2
Word 3, Reading 3, Part of Speech 3

Example of creation


__________2020-04 -09_18.35.57.png

  • Enter the words from the first line, not including the header line.
  • Duplicate character strings cannot be entered in the "Reading" field.
    Word: Sato, read: Satou
    Word: Sugar, read: Satou
  • .
  • Registering words with short readings (e.g. "tooth") may increase the number of false positives. Whenever possible, please register words with a reading of three or more characters.
  • The following five types of parts of speech can be registered.
    (General nouns, addresses, company names, product/service names, and people's names)
  • Only character strings consisting of the following tokens can be registered in the "Reading" field.

How to import in bulk

  1. In MiiTel Analytics [Settings] __________2019-10-01_18.33.51.png.
  2. Click [User dictionary for speech recognition].
  3. Click [Import] button.


  1. Click the Select File button and select the CSV file to import.


  1. Select the character encoding used for the CSV file to be imported.


  1. Set the separator used in the CSV file to import.


  1. After the above settings, click the "Import" button.


Was this article helpful?