enteprise bot updates to include chitchat in qna (#1218)

Author: lauren-mills

articles/v4sdk/bot-builder-enterprise-template-overview-detail.md

@@ -26,9 +26,9 @@ A simple introduction card is provided as standard which you can adapt as needed
 
 Every Bot should handle a base level of conversational language understanding. Greetings for example are a basic thing every Bot should handle with ease. Typically, developers need to create these base intents and provide initial training data to get started. The Enterprise Bot Template provides example LU files to get you started and avoids every project having to create these each time and ensures a base level of capability out of the box.
 
-The LU files provide the following intents across English, French, Italian, German and Spanish.
+The LU files provide the following intents across English, Chinese, French, Italian, German, Spanish.
 
-> Greeting, Help, Cancel, Restart, Escalate, ConfirmYes, ConfirmNo, ConfirmMore, Next, Goodbye
+> Cancel, Confirm, Escalate, FinishTask, GoBack, Help, Reject, Repeat, SelectAny, SelectItem, SelectNone, ShowNext, ShowPrevious, StartOver, Stop
 
 The [LU](https://github.com/Microsoft/botbuilder-tools/blob/master/packages/Ludown/docs/lu-file-format.md) format is similar to MarkDown enabling easy modification and source control. The [LuDown](https://github.com/Microsoft/botbuilder-tools/tree/master/packages/Ludown) tool is then used to convert .LU files into LUIS models which can then be published to your LUIS subscription either through the portal or the associated [LUIS](https://github.com/Microsoft/botbuilder-tools/tree/master/packages/LUIS) CLI (command line) tool.
 
@@ -40,37 +40,37 @@ A middleware component is provided that screen texts and surfaces through a ```T
 
 ## Telemetry
 
-Providing insights into the user engagement of your Bot has proven to be highly valuable. This insight can help you understand the levels of user engagement, what features of the Bot they are using (intents) along with questions people are asking that the Bot isn't able to answer - highlighting gaps in the Bot's knowledge that could be addressed through new QnAMaker articles for instance.
+Providing insights into the user engagement of your Bot has proven to be highly valuable. This insight can help you understand the levels of user engagement, what features of the Bot they are using (intents) along with questions people are asking that the Bot isn't able to answer - highlighting gaps in the Bot's knowledge that could be addressed through new QnA Maker articles for instance.
 
-Integration of Application Insights provides significant operational/technical insight out of the box but this can also be used to capture specific Bot related events - messages sent and recieved along with LUIS and QnAMaker operations.
+Integration of Application Insights provides significant operational/technical insight out of the box but this can also be used to capture specific Bot related events - messages sent and recieved along with LUIS and QnA Maker operations.
 
 Bot level telemetry is intrinsically linked to technical and operational telemetry enabling you to inspect how a given user question was answered and vice versa.
 
-A middleware component combined with a wrapper class around the QnAMaker and LuisRecognizer SDK classes provides an elegant way to collect a consistent set of events. These consistent events can then be used by the Applicatin Insights tooling along with tools like PowerBI.
+A middleware component combined with a wrapper class around the QnA Maker and LuisRecognizer SDK classes provides an elegant way to collect a consistent set of events. These consistent events can then be used by the Applicatin Insights tooling along with tools like PowerBI.
 
 An example PowerBI dashboard is provided with each project created using the Enterprise Bot Template. See the [PowerBI](bot-builder-enterprise-template-powerbi.md) section for more information.
 
 ## Dispatcher
 
-A key design pattern used to good effect in the first wave of covnersational experiences was to leverage Language Understanding (LUIS) and QnAMaker. LUIS would be trained with tasks that your Bot could do for an end user and QnAMaker would be trained with more general knowledge.
+A key design pattern used to good effect in the first wave of covnersational experiences was to leverage Language Understanding (LUIS) and QnA Maker. LUIS would be trained with tasks that your Bot could do for an end user and QnA Maker would be trained with more general knowledge.
 
-All incoming utterances (questions) would be routed to LUIS for analysis. If the intent of a given utterance was not identified it was marked as a None intent. QnAMaker was then used to try and find an answer for the end-user.
+All incoming utterances (questions) would be routed to LUIS for analysis. If the intent of a given utterance was not identified it was marked as a None intent. QnA Maker was then used to try and find an answer for the end-user.
 
 Whilst this pattern worked well there were two key scenarios where problems could be experienced.
 
-- If  utterances in the LUIS model and QnAMaker overlapped sometimes slightly, this could lead to strange behavior where LUIS may try to process a question when it should have been directed to QnaMaker.
+- If  utterances in the LUIS model and QnA Maker overlapped sometimes slightly, this could lead to strange behavior where LUIS may try to process a question when it should have been directed to QnA Maker.
 - When there were two or more LUIS models a Bot would have to invoke each one and perform some form of  intent evaluation comparison to identify where to send a given utterance. As there is no common baseline score comparison across models didn't work effectively leading to a poor user experience.
 
-The [Dispatcher](https://docs.microsoft.com/en-us/azure/bot-service/bot-builder-tutorial-dispatch?view=azure-bot-service-4.0&tabs=csaddref%2Ccsbotconfig) provides an elegant solution to this by extracting utterances from each configured LUIS model and questions from QnAMaker and creating a central dispatch LUIS model.
+The [Dispatcher](https://docs.microsoft.com/en-us/azure/bot-service/bot-builder-tutorial-dispatch?view=azure-bot-service-4.0&tabs=csaddref%2Ccsbotconfig) provides an elegant solution to this by extracting utterances from each configured LUIS model and questions from QnA Maker and creating a central dispatch LUIS model.
 
-This enables a Bot to quickly identify which LUIS model or component should handle a given utterance and ensures QnAMaker data is considered at the top level of intent processing not just the None intent as before.
+This enables a Bot to quickly identify which LUIS model or component should handle a given utterance and ensures QnA Maker data is considered at the top level of intent processing not just the None intent as before.
 
-This Dispatch tool also enables evaluation which will highlight confusion and overlap across LUIS models and QnAMaker knowledgebases highlighting issues before deployment.
+This Dispatch tool also enables evaluation which will highlight confusion and overlap across LUIS models and QnA Maker knowledgebases highlighting issues before deployment.
 
 The Dispatcher is used at the core of each project created using the Enterprise Bot Template. The Dispatch model is used within the `MainDialog` class to identify whether the target is a LUIS model or QnA. In the case of LUIS, the secondary LUIS model is invoked returning the intent and entities as usual.
 
-## QnAMaker
+## QnA Maker
 
-[QnAMaker](https://www.qnamaker.ai/) provides the ability for non-developers to curate general knowledge in the format of question and answer pairs. This knowledge can be imported from FAQ data sources, product manuals and interactively within the QnaMaker portal.
+[QnA Maker](https://www.qnamaker.ai/) provides the ability for non-developers to curate general knowledge in the format of question and answer pairs. This knowledge can be imported from FAQ data sources, product manuals and interactively within the QnaMaker portal.
 
-An example set of QnA entries is provided in the [LU](https://github.com/Microsoft/botbuilder-tools/blob/master/packages/Ludown/docs/lu-file-format.md) file format within the QnA folder of CogSvcModels. [LuDown](https://github.com/Microsoft/botbuilder-tools/tree/master/packages/Ludown) is then used as part of the deployment script to create a QnAMaker JSON file which the [QnAMaker](https://github.com/Microsoft/botbuilder-tools/tree/master/packages/QnAMaker) CLI (command line) tool then uses to publish items to the QnAMaker knowledgebase.
+Two example QnA Maker models are provided in the [LU](https://github.com/Microsoft/botbuilder-tools/blob/master/packages/Ludown/docs/lu-file-format.md) file format within the QnA folder of CognitiveModels, one for FAQ and one for chit-chat. [LuDown](https://github.com/Microsoft/botbuilder-tools/tree/master/packages/Ludown) is then used as part of the deployment script to create a QnA Maker JSON file which the [QnA Maker](https://github.com/Microsoft/botbuilder-tools/tree/master/packages/QnAMaker) CLI (command line) tool then uses to publish items to the QnA Maker knowledgebase.

articles/v4sdk/bot-builder-enterprise-template-overview.md

@@ -28,7 +28,7 @@ Basic conversational responses  | Responses to basic conversational intents abst
 Inappropriate content or PII (personally identifiable information) detection  |Detect inappropriate or PII data in incoming conversations through use of [Content Moderator](https://azure.microsoft.com/en-us/services/cognitive-services/content-moderator/) in a middleware component.
 Transcripts  | Transcripts of all conversation stored in Azure Storage
 Dispatcher | An integrated [Dispatch](https://docs.microsoft.com/en-us/azure/bot-service/bot-builder-tutorial-dispatch?view=azure-bot-service-4.0&tabs=csaddref%2Ccsbotconfig) model to identify whether a given utterance should be processed by LUIS + Code or passed to QnA Maker.
-QnA Maker Integration  | Integration with [QnA Maker](https://www.qnamaker.ai) to answer general questions from a Knowledgebase which can be leverage existing data sources (e.g. PDF manuals).
+QnA Maker Integration  | Integration with [QnA Maker](https://www.qnamaker.ai) to answer general questions from a Knowledgebase which can be leverage existing data sources (e.g. PDF manuals). A QnA Maker chit-chat model is also included to provide standard answers to common queries ([learn more](https://docs.microsoft.com/en-us/azure/cognitive-services/qnamaker/how-to/chit-chat-knowledge-base)).
 Conversational Insights  | Integration with [Application Insights](https://azure.microsoft.com/en-gb/services/application-insights/) to collect telemetry for all conversations and an example PowerBI dashboard to get you started with insights into your conversational experiences.
 
 In addition, all of the Azure resources required for the Bot are automatically deployed: Bot registration, Azure App Service, LUIS, QnA Maker, Content Moderator, CosmosDB, Azure Storage, and Application Insights. Additionally, base LUIS, QnA Maker, and Dispatch models are created, trained, and published to enable immediate testing of basic intents and routing.