generic-dialog

• 1: generic-dialog v1
• 2: generic-dialog-v3

1 - generic-dialog v1

Aura Bot generic-dialog v1

Introduction

The generic-dialog allows OBs to implement certain local use cases for Movistar Home, Movistar Plus and Set-top Box (STB) channels in an autonomous, agile and simple way with no code development, only by means of configuration.

Use this generic-dialog if your channelData version is v1.

You can check its source code in Github: generic-dialog.ts.

Included intents

The next table shows current intents available on generic-dialog and the previous library where the intent was set.

Generic indexed text

Generic dialog implements an algorithm, which based on a dialog configuration, can customize its response. The getIndexedText function returns a POEditor resource based on the dialog configuration’s setting indexedResourceBase and the value entity.entity obtained from the first entity. This functionality relies in those two values and will return an error if they are not present.

{
    "intent": "intent.exact-match.faq",
    "settings": {
        "indexedResourceBase": "generic:faq-answer.",
        "functions": {
            "params": [
                "context",
                "currentIntent",
                "persistentData"
            ],
            "algorithm": "getIndexedText"
        }
    }
}

In this example, the reference that will be searched in POEditor will be generic:faq-answer.X, where X is the identifier gathered from the first entity.

Keep in mind that in function.params there has to be an array of parameters that will be used in getIndexedText.

Enabling AI resolution

To enable the AI-based fallback mechanism, add the flag enableAIResolution to true inside the dialog configuration for the intent. This setting allows the dialog to use the ai recognizer when the NLP resolution returns no entities or when the Movistar Plus resolution returns no results.

Example configuration:

{
    "intent": "intent.tv.search",
    "settings": {
        "locales": {
            "error": [
                "core:flc.answer.error"
            ]
        },
        "enableAIResolution": true,
        "needTvResolution": true
    }
}

AI fallback resolution

When using generic-dialog v1 and needTvResolution is set to true, a fallback mechanism using AI services is available if enableAIResolution is set to true in the dialog configuration.

If NLP resolution returns no entities or when the Movistar Plus resolution returns no results, the system applies an AI-based fallback strategy (“triage”). This ensures robustness and continuity of the dialog flow.

The fallback is triggered in two scenarios:

1. No entities detected If the intent requires TV resolution and no entities are returned by the NLP (intentResult.entities?.length === 0), the dialog calls the recognizeUsingAI function. If a new intent is successfully recognized, the current dialog is replaced by the new one.

2. Movistar Plus resolution returns no results When the intent and entities are passed to Movistar Plus Resolution but no valid result is obtained (GENERIC_MP_RESOLUTION_NO_RESULT_CODE), the system again invokes recognizeUsingAI. A new intent will be attempted and, if available, the dialog is redirected accordingly.

In both cases, the interaction origin is marked as InteractionOriginType.NoResults.

2 - generic-dialog-v3

Aura Bot generic-dialog-v3

Introduction

The generic-dialog-v3 allows OBs to implement certain local use cases for channelData v3 channels in an autonomous, agile and simple way with no code development, only by means of configuration.

You can check its source code in Github: generic-v3-dialog.ts.

Differences between generic dialog v1 and v3

These are the main differences in comparison with the generic-dialog v1:

• Deprecated fields in dialog’s settings (dialogLibraries.dialogs.triggerConditions.settings):

  • type, action, payload, addEntitiesToPayload, sound, payloadType and isFullScreen are no longer used.
  • action, payload and sound configurations will be transformed to the new actions format and added to the new field actions.

• New field actions: This field contains the configuration of all the actions returned inside the channelData.

Generic dialog environment variables

Generic

| TV_TIMER_RECOGNIZER_SECOND_ENTITY_CONFIG | string | Local | Configuration of recognition of entities with timer content in second position to calculate number of seconds. By default: 1 4:0.25,3 4s:0.75,4 de hora:0.25,hora:1.0,media:0.5,medio:0.5,minuto:1.0,un:1.0,un 4:0.25,una:1.0,4:0.25. |

Generic dialog utilities

There are several functions that implement different requirements for the generic-dialog such as:

• calculateSecondLevels: Calculate the value in seconds in case the entity type is not expressed in this format: ent.time_length_sec. Right now it only converts minutes into seconds.
• getIndexedText: Get the text’s reference with information about the entity value.
• getInfoWifiPlus: Retrieve Wi-Fi info.
• getSecondsLevelsWithTimerRecognizer: Calculates the total number of seconds, based on the provided timer entities, making a recognition and conversion of them depending on the configuration of each country. The configuration is added in the environment variables: TV_TIMER_RECOGNIZER_FIRST_ENTITY_CONFIG and TV_TIMER_RECOGNIZER_SECOND_ENTITY_CONFIG.
• getWifiInfo: Manage Wi-Fi data by setting the persistentData and the value of movistarPlusResolution.

AI fallback resolution

The configuration and fallback mechanism for AI resolution in generic-dialog-v3 are the same as in generic-dialog v1.

The only difference is in the second fallback condition:

• In generic-dialog v1, the fallback is triggered when Movistar Plus resolution returns no results.
• In generic-dialog v3, the fallback is triggered when the complex logic resolution process returns no results.

All other aspects (configuration with enableAIResolution, marking the interaction origin as InteractionOriginType.NoResults, and redirecting to the new dialog) remain exactly the same.