亚马逊 Alexa 技能

Posted 2023-04-14

【中文标题】亚马逊 Alexa 技能

【英文标题】：Amazon Alexa Skill

【发布时间】：2019-02-04 05:28:53

【问题描述】：

我想问 alexa 不同类型的问题，最后我希望它应该问“你还有什么想知道的吗？”当我说是时（是的工作建议），它应该根据我的意图向我提出建议。就像我在

IncityIntent:

    'InCityIntent': function () 
        speechOutput = '';


speechOutput = "The atmosphere in the city is beautiful. Is there anything else you would like to know";
        this.emit(":ask", speechOutput, speechOutput);


'YesIntent': function () 
        speechOutput = '';
/*when the user say yes, he should get this output*/  
            speechOutput = You can learn more about city by trying, alexa what are the best places in the city";
            this.emit(":tell",speechOutput, speechOutput);

FoodIntent:

    'FoodIntent': function () 
        speechOutput = '';


speechOutput = "Food in the city is delicious. Is there anything else you would like to know";
        this.emit(":ask", speechOutput, speechOutput);

'YesIntent': function () 
        speechOutput = '';
/*change in response here*/
            speechOutput = You can learn more about food by trying, alexa what are the best restaurants in the city";
            this.emit(":tell",speechOutput, speechOutput);

【参考方案1】：

首先，不要创建自定义YesIntent 和NoIntent，而是使用AMAZON.YesIntent 和AMAZON.NoIntent。如果您愿意，您可以随时向这些预定义的意图添加话语。

您的问题可以通过多种方式解决。

使用 sessionAttributes： 当您收到初始请求时，添加previousIntent 属性或其他内容以跟踪sessionAttributes 中的转换，例如InCityIntent。在您的 AMAZON.YesIntent 或 AMAZON.NoIntent 处理程序中检查先前的意图并做出相应的回复。

 'InCityIntent': function () 
       const speechOutput = "The atmosphere in the city is beautiful. Is there anything else you would like to know";
       const reprompt = "Is there anything else you would like to know";
       this.attributes['previousIntent'] = "InCityIntent";
       this.emit(":ask", speechOutput, reprompt);
  

 'Amazon.YesIntent': function () 
     var speechOutput =  "";
     var reprompt = "";
     if (this.attributes 
        && this.attributes.previousIntent
        && this.attributes.previousIntent === 'InCityIntent' ) 
        speechOutput = "You can learn more about city by trying, Alexa what are the best places in the city";
        reprompt = "your reprompt";
      else if ( //check for FoodIntent ) 

       // do accordingly
     
     this.attributes['previousIntent'] = "Amazon.YesIntent";
     this.emit(":ask", speechOutput, reprompt);
  

使用状态处理程序ask-nodejs-sdk v1 具有状态处理程序来根据状态生成响应。思路类似，sdk 会为你添加sessionAttribute 参数，when 会自动将处理程序映射到该状态。

 'InCityIntent': function () 
       const speechOutput = "The atmosphere in the city is beautiful. Is there anything else you would like to know";
       const reprompt = "Is there anything else you would like to know";
       this.handler.state = "ANYTHING_ELSE";
       this.emit(":ask", speechOutput, reprompt);
  

const stateHandlers = Alexa.CreateStateHandler("ANYTHING_ELSE", 
    "AMAZON.YesIntent": function () 
       var speechOutput =  "You can learn more about city by trying, Alexa what are the best places in the city";
       var reprompt = "your reprompt";
       this.emit(":ask", speechOutput, reprompt);
    ,

一旦设置了state，下一次将触发在该特定状态处理程序中定义的意图处理程序。相应地更改您的state，完成后将其删除。

【讨论】：

感谢您的回复，但以上方法均无效。它给出了没有响应的错误。

请勿复制粘贴代码，仅供参考。

我没有复制粘贴。当我尝试运行此代码时，如果我用我的技能名称说是，比如告诉 citytravel 是，那么它会从我第一次声明是的意图重复对话，如果我只是说是，那么它会说存在问题请求技能的响应。

进入技能会话后，您不必每次都使用技能调用名称。