Skip to content

Latest commit

聽

History

History
334 lines (245 loc) 路 8.93 KB

slots.rst

File metadata and controls

334 lines (245 loc) 路 8.93 KB
Raw
desc:Store information the user provided as well as information from database queries in slots to influence how the machine learning based dialogue continues.

Slots

What are slots?

Slots are your bot's memory. They act as a key-value store which can be used to store information the user provided (e.g their home city) as well as information gathered about the outside world (e.g. the result of a database query).

Most of the time, you want slots to influence how the dialogue progresses. There are different slot types for different behaviors.

For example, if your user has provided their home city, you might have a text slot called home_city. If the user asks for the weather, and you don't know their home city, you will have to ask them for it. A text slot only tells Rasa Core whether the slot has a value. The specific value of a text slot (e.g. Bangalore or New York or Hong Kong) doesn't make any difference.

If the value itself is important, use a categorical or a bool slot. There are also float, and list slots. If you just want to store some data, but don't want it to affect the flow of the conversation, use an unfeaturized slot.

How Rasa Uses Slots

The Policy doesn't have access to the value of your slots. It receives a featurized representation. As mentioned above, for a text slot the value is irrelevant. The policy just sees a 1 or 0 depending on whether it is set.

You should choose your slot types carefully!

How Slots Get Set

You can provide an initial value for a slot in your domain file:

slots:
  name:
    type: text
    initial_value: "human"

There are multiple ways that slots are set during a conversation:

Slots Set from NLU

If your NLU model picks up an entity, and your domain contains a slot with the same name, the slot will be set automatically. For example:

# story_01
* greet{"name": "Ali"}
  - slot{"name": "Ali"}
  - utter_greet

In this case, you don't have to include the - slot{} part in the story, because it is automatically picked up.

Slots Set By Clicking Buttons

You can use buttons as a shortcut. Rasa Core will send messages starting with a / to the RegexInterpreter, which expects NLU input in the same format as in story files, e.g. /intent{entities}. For example, if you let users choose a color by clicking a button, the button payloads might be /choose{"color": "blue"} and /choose{"color": "red"}.

You can specify this in your domain file like this: (see details in :ref:`domains`)

utter_ask_color:
- text: "what color would you like?"
  buttons:
  - title: "blue"
    payload: '/choose{"color": "blue"}'
  - title: "red"
    payload: '/choose{"color": "red"}'

Slots Set by Actions

The second option is to set slots by returning events in :ref:`custom actions <custom-actions>`. In this case, your stories need to include the slots. For example, you have a custom action to fetch a user's profile, and you have a categorical slot called account_type. When the fetch_profile action is run, it returns a :class:`rasa.core.events.SlotSet` event:

slots:
   account_type:
      type: categorical
      values:
      - premium
      - basic
from rasa_sdk.actions import Action
from rasa_sdk.events import SlotSet
import requests

class FetchProfileAction(Action):
    def name(self):
        return "fetch_profile"

    def run(self, dispatcher, tracker, domain):
        url = "http://myprofileurl.com"
        data = requests.get(url).json
        return [SlotSet("account_type", data["account_type"])]
# story_01
* greet
  - action_fetch_profile
  - slot{"account_type" : "premium"}
  - utter_welcome_premium

# story_02
* greet
  - action_fetch_profile
  - slot{"account_type" : "basic"}
  - utter_welcome_basic

In this case you do have to include the - slot{} part in your stories. Rasa Core will learn to use this information to decide on the correct action to take (in this case, utter_welcome_premium or utter_welcome_basic).

Note

It is very easy to forget about slots if you are writing stories by hand. We strongly recommend that you build up these stories using :ref:`section_interactive_learning_forms` rather than writing them.

Slot Types

Text Slot

.. option:: text

  :Use For: User preferences where you only care whether or not they've
            been specified.
  :Example:
     .. sourcecode:: yaml

        slots:
           cuisine:
              type: text
  :Description:
      Results in the feature of the slot being set to ``1`` if any value is set.
      Otherwise the feature will be set to ``0`` (no value is set).

Boolean Slot

.. option:: bool

  :Use For: True or False
  :Example:
     .. sourcecode:: yaml

        slots:
           is_authenticated:
              type: bool
  :Description:
      Checks if slot is set and if True

Categorical Slot

.. option:: categorical

  :Use For: Slots which can take one of N values
  :Example:
     .. sourcecode:: yaml

        slots:
           risk_level:
              type: categorical
              values:
              - low
              - medium
              - high

  :Description:
     Creates a one-hot encoding describing which of the ``values`` matched.

Float Slot

.. option:: float

  :Use For: Continuous values
  :Example:
     .. sourcecode:: yaml

        slots:
           temperature:
              type: float
              min_value: -100.0
              max_value:  100.0

  :Defaults: ``max_value=1.0``, ``min_value=0.0``
  :Description:
     All values below ``min_value`` will be treated as ``min_value``, the same
     happens for values above ``max_value``. Hence, if ``max_value`` is set to
     ``1``, there is no difference between the slot values ``2`` and ``3.5`` in
     terms of featurization (e.g. both values will influence the dialogue in
     the same way and the model can not learn to differentiate between them).

List Slot

.. option:: list

  :Use For: Lists of values
  :Example:
     .. sourcecode:: yaml

        slots:
           shopping_items:
              type: list
  :Description:
      The feature of this slot is set to ``1`` if a value with a list is set,
      where the list is not empty. If no value is set, or the empty list is the
      set value, the feature will be ``0``. The **length of the list stored in
      the slot does not influence the dialogue**.

Unfeaturized Slot

.. option:: unfeaturized

  :Use For: Data you want to store which shouldn't influence the dialogue flow
  :Example:
     .. sourcecode:: yaml

        slots:
           internal_user_id:
              type: unfeaturized
  :Description:
      There will not be any featurization of this slot, hence its value does
      not influence the dialogue flow and is ignored when predicting the next
      action the bot should run.

Custom Slot Types

Maybe your restaurant booking system can only handle bookings for up to 6 people. In this case you want the value of the slot to influence the next selected action (and not just whether it's been specified). You can do this by defining a custom slot class.

In the code below, we define a slot class called NumberOfPeopleSlot. The featurization defines how the value of this slot gets converted to a vector to our machine learning model can deal with. Our slot has three possible "values", which we can represent with a vector of length 2.

(0,0) not yet set
(1,0) between 1 and 6
(0,1) more than 6 
.. testcode::

   from rasa.core.slots import Slot

   class NumberOfPeopleSlot(Slot):

       def feature_dimensionality(self):
           return 2

       def as_feature(self):
           r = [0.0] * self.feature_dimensionality()
           if self.value:
               if self.value <= 6:
                   r[0] = 1.0
               else:
                   r[1] = 1.0
       return r

Now we also need some training stories, so that Rasa Core can learn from these how to handle the different situations:

# story1
...
* inform{"people": "3"}
  - action_book_table
...
# story2
* inform{"people": "9"}
  - action_explain_table_limit