Adding dictionary-based entities is for entities for which you define terms, synonyms or Regex. When the Virtual Assistant is prompted, it will recognize only those user-inputs that exactly match the synonyms or Regex that you define.
Synonyms
With the synonym entity, you can define categorize a bunch of terms as one entity, say “vehicles” and define synonyms that come under it by providing examples, like “car”, “bike”, “truck”, et cetera.
For example, you can define an entity named “Vehicles” and define many synonyms under it such as cars, bikes, trucks, etc. Depending on your use-case, you can go about defining the main entity and define many sub-class terms in the form of synonyms. This helps the Virtual Assistant understand what the user is intending to prompt based on the main entity category.
Regex
With the Regex entity, you can define a category of terms as an entity and add one or more values in that entity. For example, your entity category can be “contact_details” and the term you define under it can be “phone_number”. With a Regex entity, you can set the exact length of the phone number that your users will prompt the virtual assistant with. For example, in India, a mobile number is ten characters in length and you can set the Regex with a regular expression like: “(\d{10})”. Setting this regular expression will help your virtual assistant to prompt the user to enter a correct number if the character length is too short, or too big.
The Regex field allows you to define specific sequences for a particular entity value. A Regex in its default must be entered as a regular expression in the text field.
- For each entity that you create, you can have up to five Regex.
- Each Regex that you create is limited to 512 characters in length.
Consider a normal USA phone number for example:
You can add a regular expression inside a Regex entity to help the Virtual Assistant recognize phone numbers that you want it to recognize. This will help the Virtual Assistant distinguish relevant phone numbers from incorrect or irrelevant ones.
You can create a “contact_info” entity that can store the phone numbers of users. To allow the Virtual Assistant to recognize a proper phone number, you can enter a regular expression. The regular expression in this case can be: “^(1\s?)?((\([0-9]{3}\))|[0-9]{3})[\s\-]?[\0-9]{3}[\s\-]?[0-9]{4}$ ” which will recognize general US phone numbers.
Once you’re done adding a Regex or synonym, you can click on Save and you’re all set.
The entity that you have created will be stored in the entities tab and will be accessible at any time if you choose to either make changes to it or delete it altogether.