/
Classifier

Classifier

Owned by Olesya Tantsyura
Last updated: Dec 19, 2024
2 min read

1. Description

A cluster is a classifier of subscriber response types.

This component is present in flow types like:

  • Voice;

  • Chat.

The block (Fig.1(1)) is used in space to build the Flow.  The settings panel for this component opens (Fig.1(2)) when you click on the block.

Fig.1. Classifier component

1.1. Classifier block

The block consists of the following elements:

  1. The name of the component;

  2. In branch — receiving (connecting with the previous block) branch;

  3. Out branch — outgoing (connecting with the following block) branch. When hovering over the Out field, a switch appears. The switch allows this component to be connected to a component that already has a connection.

The presence of an icon indicates that the switch is on, and this component can be connected to a component that already has a connection.

Changing the switch position removes the outgoing branch of this component.

The button deletes the block from the Flow Schema.

1.2. Classifier settings panel

It consists of the following elements:

  1. Input;

  2. Phrase search;

  3. Set;

  4. Clusters;

  5. Match type;

  6. Break;

  7. Limit;

  8. Add description.

1.2.1. Input

Here, enter what needs to be passed to the input. For example, the variable that was set in the Recv Message component.

1.2.2. Phrase search

Responsible for determining how the search will occur:

- search by word;

— search by phrase.

1.2.3. Set

The variable's name is indicated, to which the data received from the clusters are transferred.

1.2.4. Clusters

It is designed to create a cluster (Fig. 2).

When you click on the Add button, the following options are displayed:

  • Key — cluster key;

  • Value —  enter the list of possible words/phrases by which it determines whether a message belongs to this cluster;

  • Delete  — used to delete a cluster.

Fig. 2. Clusters

Using the Switch component, you can create branches along which the circuit will pass according to the distribution by clusters.

To do this:
—  in the Cases field of the Switch component, you have to enter the cluster keys created in the Classifier component;
— in the Variable field of the Switch component — the variable whose name is entered in the Set field of the Classifier component.

1.2.5. Match type

Responsible for searching for a keyword by full or partial match.

There are two options:

  • full — keyword search by full match;

  • part — keyword search by partial match.

1.2.6. Break

1.2.7. Add limit

Fig. 2. Setting a limit

1.2.8. Add description

Fig. 3. Add description