Answer Question
Verb: answerQuestion
Available from: <Enterprise>
Answers questions, based on the answers and questions contained in the Knowledge Base used for the interaction.
Syntax
answerQuestion --kb(String) [--version(Numeric)] [--minimumscore(Numeric)] [--topn(Numeric)] [--options(Nullable<AnswerQuestionOptions>)] [--botHistoryMessageId(String)] --culture(Culture) --text(String) (String)=answer (String)=tags (String)=context (Numeric)=score (List<String>)=options (DataTable)=data (Boolean)=success
Inputs
| Script | Designer | Required | AcceptedTypes | Description |
|---|---|---|---|---|
| --kb | Knowledge Base | Required | Text | Knowledge Base that will be used to query the answer based on the user's question.
For the execution of the command to occur, it is necessary to first create a Knowledge Base, in which it must contain the possible questions from the user and the closest answers to the question asked. This base is built according to a template and used on IBM Robotic Process Automation Studio, through the menu "Tools> Model Builder Machine Learning". |
| --version | Version | Optional | Number | Knowledge Base version used. |
| --context | Context(Obsolete) | Optional | Text | Context to look for on the analysed text.
1. The context provides patterns and themes that the Chatbot should look for to give more accurate answers to the user. 2. This parameter is obsolete, the output parameter Context should be used instead. |
| --minimumscore | Minimum Score | Optional | Number | Minimum acceptable score of the answer in relation to the user's question. |
| --topn | Answers Quantity | Optional | Number | Number of responses that are returned, which may bring more than one answer from the Knowledge Base used. |
| --options | Options | Optional | AnswerQuestionOptions | Option that will be used if, when querying answers in the Knowledge Base, other forms of research might be considered as well, such as:
|
| --botHistoryMessageId | Bot Ask History | Optional | Text | Identifier referring to the conversation history returned by the commands of the category "Communication" and subcategory "Ask", and which will associate the responses returned by the Answer Question command to the history. |
| --culture | Language | Required | Culture | Language that is used in the bot conversation:
The variable of type "Language" can be obtained previously through the Create a Language command. |
| --text | Text | Required | Text | Text that contains the user's response. |
Outputs
| Script | Designer | AcceptedTypes | Description |
|---|---|---|---|
| answer | Answer | Text | Returns the answer that has the highest score in relation to the user's question. |
| tags | Tags | Text | Returns the tags for the context of the question and answer in the interaction.
The labels are configured in the Knowledge Base used by the command. |
| context | Context | Text | Returns the context of the conversation. |
| score | Score | Number | Returns the score for the best answer. |
| options | Additional Answers | List<Text> | Returns a list of additional responses to the user's question. |
| data | Data | Data Table | Returns a data table with information about the questions, answers, contexts, scores and labels. |
| success | Success | Boolean | Returns "True", if the script runs successfully, or "False", otherwise. |
Example
In the following example, the Answer Question command takes the user's speech, returned by the Bot Ask command to analyze the text based on the "KB-flower-shop" knowledge base model. Then, it returns the most appropriate answer to the user's question. In this case, the user is able to ask questions such as "What kind of products do you have?" or "What chocolates do you have?".
defVar --name userResponse --type String
defVar --name betterResponse --type String
defVar --name tag --type String
defVar --name botLanguage --type Language
createLanguage --culture en-US botLanguage=value
botConnect --type "Chat" --language ${botLanguage} --autoanswer --timeout 00:05:00
botAsk --retry --attempts 3 --retrytext "What is your question?" --language ${botLanguage} --text "Hello! How can I help you?" --timeout 00:01:00 userResponse=value
while --left "${tag}" --operator "Equal_To" --right closure --negate
answerQuestion --kb "KB-flower-shop" --minimumscore 500 --culture en-US --text "${userResponse}" betterResponse=answer tag=tags
if --left "${tag}" --operator "Equal_To" --right "+closure"
botSay --text "${betterResponse}"
break
endIf
botAsk --retry --attempts 3 --retrytext "What is your question?" --language ${botLanguage} --text "${betterResponse}" --timeout 00:01:00 userResponse=value
endWhile
botDisconnect
For the above script to work correctly, the knowledge base must be published using the Machine Learning Model Builder in IBM RPA Studio.