I have created a chatbot with IBM Waston Assistant with several questions having binary answers(Yes, No).
Eg: Would you like to share your feedback with us? (Yes/ No)
Are you comfortable with our environment (Yes/ No)
I need to recognize correct entity value(yes/no) for each question. Is there any possible way?
I see two possible ways to go:
1. Create two intents, e.g., "positive_answer" and "negative_answer" and provide examples like "yes", "yeah", "definitely", "I feel ok", "Yes, I want", etc. to train them (use "no", "I don't want to", etc. for the negative part). In your dialog tree, after posing the question react to the positive or negative answer appropriately.
2. Do the same, but define entities like "positive_answer_word" or "negative_answer_word" and train and react accordingly.
You can reuse the same intent or entity for the different questions. Key is to react accordingly in a child node of the dialog tree.
You can do this with a single entity. Each question is separate, so you can just reuse that same entity for each one, but save it as a different context variable. Then just like in the tutorial linked above, you will save that same entity to multiple context variables, and when the session ends you can write the whole thing to your database.
Related
I am gathering some context variables with slots, and they work just fine.
So I decided to do in another node of the conversation, check if one of these context variables is a specific number:
I was thinking on enabling multi-responses and check if, for example $dni:1 (it is an integer, pattern of 1 integer only), or if it is 2 or 3:
But this is not working. I was trying to solve it for some days with different approaches but I really cannot find a way through it.
My guess is that a context variable has a value, and you can print it to use it like responding with the user's name and stuff like that (which indeed is useful!), but comparing values is not possible.
Any insights on this I can receive?
Watson Assistant uses a short-hand syntax but also supports the more complex expressions. What you could do is to edit the condition in the JSON editor. There, for the condition, use a function like matches() on the value of the context variable.
Note that it is not recommended to check for context variables in the slot conditions. You can use multi-responses. An alternative way is to put the check into the response itself. There, you can use predicates to generate the answer.
<? context.dni==1 ? 'Very well' : 'Your number is not 1' ?>
You can nest the evaluation to have three different answers. Another way is to build an array of responses and use dni as key.
Instead of matching to specific integers, you could consider using the Numbers system entity. Watson Assistant supports several languages. As a benefit, users could answer "the first one", "the 2nd option", etc., and the bot still would understand and your logic could still route to the correct answer.
I would like to build an app with a oral code verification.
i could just set my cde in dialogflow before then, juste verify it.
GH : "For continue, give me the code"
Me : " 1 2 3 4"
GH " Access granted" / "Access denied"
But how can do an input a get this code on dialogflow?
First of all - consider if you really want to do this. Having someone say a passcode out loud isn't really very secure and adds very little additional security in a multi-user environment.
There are two stages to this - the first is setting up an Intent to handle this, specifically in the format you want, and the second would be handling and verifying this is the correct code.
Setting up the Intent
We'll need two intents - one that prompts and sets a context so we know we're expecting the validation code, and one that checks for the code.
The prompting intent might look something like this:
The notable part here is that it is setting an output context. We'll see why that matters in a moment.
The one to handle numeric input might look like this:
There is a lot more to this one. First note that we're requiring an input context that matches the output context from the last Intent. This means that this Intent should only match if that Context has been set. This lets us talk about numbers elsewhere in our conversation without triggering this validation.
Next we're looking for sequences of numbers that match the #sys.number-sequence built-in Entity type. There are other entity types that may be useful for you - see the documentation for details and pick one that makes sense or experiment to find what works best in your case.
Finally, we're going to use a webhook for fulfillment to verify if the code is correct. Which is the next session...
Verifying the code
While there are ways to do the verification without a webhook, this is really the most straightforward way to do it. If you're using Google's library to handle input from Dialogflow, you can get the value with something like
var code = app.getArgument('number-sequence');
using whatever the parameter name is. If you're not using the library, you can find this in the JSON at result.parameters.number-sequence.
You would then verify this code, however you want, and return a message indicating if it is correct or not.
If you want to use a sequence of numbers as your code you can use the #sys.number-sequence entity to recognize it and then check the code in your webhook.
Another way would be to simply make a custom entity 'code' that has an entry of '1234'.
I'm trying to create a chat bot that will help users search up motorcycles.
I'm new to API.AI and have set up my entities and their synonyms, my intent and user expressions, as well as references to the entities (#engineSize, #make, #bikeType).
My problem is when I try to add a required action and prompt, and then try to save the intent, I get the following message:
"The following entities reference each other and form an infinite loop: [engineSize]."
Initially I thought I was using the references wrong in the user expressions. I deleted every reference except for one expression which uses all three entities.
I can't figure out what I'm doing wrong. Any help would be greatly appreciated, thanks! Pix below for further details.
EDIT: I fixed one of the issues of trying to pass a template expression as an example. However, I still get the same error message. I will replace and update my image links to include the edits.
Annotated User expressions
Required Actions
Interestingly enough, the answer to this post would have been difficult to find because the problem was in defining my entities.
In the entity definitions, I included an #ref to the entity itself. ie the bikeType entity contained #bikeType as one of its definitions.
This is not to be mistaken with the User Expressions. As long as the user expression is marked as a Template (the entire line is denoted with an '#' on the far left, as opposed to a large " ), there should be no issues.
Edited for clarity to get at root problem
In the provided user input examples you give the intent, you are supposed to provide general examples and then highlight any text belonging to an entity to map where entities appear in user's inquiries.
In your case, you have input the actual entity reference '#engineSize' as an example belonging to the engineSize entity, creating a self reference.
A proper provided user example would look like:
Also note though that if you are just using entities to store generic information like numbers, addresses, times, etc. it generally makes far more sense to use prebuilt system entities for those categories than create a custom entity, for example #sys.number-integer might be exactly what you need
It looks like you need to get a firmer understanding of entities, for which I would recommend the documentation:
https://docs.api.ai/docs/concept-entities
I'm trying to use layers to make sure I separate everything into it's correct areas in my Swift / iOS / Xcode 6 project. The question, ultimately, is: is it OK / a common practice to move commonly used presentation-level logic to a separate presentation helper class instead of writing it over and over in multiple view controllers with little or no difference?
Here is an example to tie this in and give context:
One of the things I am aiming to do is use a UITableView to display report data. This UITableView will contain 7-10 rows, depending upon the user's preferences (nsuserdefaults). Each row contains a localized string for a label, and some decimal value.
As an example one row could be "Sales made this week: $500.00"
I have a reporting service class that's responsible for talking to the database and getting back / instantiating a report object. This report object contains the raw data for the report, i.e. how much you made this week, this month, this year, etc. Whether the user wants to show all these values or not is irrelevant to the service - it simply gets everything.
So since I have 3 view models that use this same report, I thought it would be wrong to rewrite the same code each time that checks the user's preferences, then creates/binds an array to the UITable and matches the labels with the report values from the object returned by the service.
A better way, I thought, was to create a presentation-level helper class whose job would be to take a report object (the thing I mentioned before that contains the report values), take a user's preferences, and then more generically combine them to create a list of localized strings matched with their respective report values, agnostic to what the view controller wants. Maybe if that requirement changes later (where different view controllers need more customization) I could use flags or different function names within that class.
This way all I have to do is something like
var report = ReportHelper.GenerateReport(reportData, userSettings)
and now report would be an object that could look like this (mock JSON data):
{"Amount made this week":"$100", "Amount made this month":"$500", "Amount made this year": "$10,000"}
And I can use this in any view controller.
The alternate is to just hard code those above values (obviously still pulling localized strings) but I don't know if I adding all those strings + checks based on user preferences + formatting. Seems more elegant to move it away.
Thanks!
I know too little about your problem to provide the definite answer, but basically you have 2 options:
- inheritance
- composition
I personally like inheritance although it is often stated you should choose composition over inheritance.
Your helper sounds somewhat like composition, so that would be the preferred setup. With your specific problem as I understand it, inheritance would lead to a duplication of data, so that is one more argument to choose composition.
So all in all you seem to be about right
Edit:
Inheritance is an "is" relationship, whereas composition is a "has" relationship.
See http://en.wikipedia.org/wiki/Object_composition for more details about composition.
And see http://en.wikipedia.org/wiki/Composition_over_inheritance for more info why and when to choose composition
I have a custom node type for which I want to have a field that uses a special combobox based on list_text. When one chooses the type list_text it is normally possible to enter a static list of selectable texts, however, I want this list to be dynamic, i.e. based on the results of a db_query. What is the best way to do this using Drupal 7?
A simple example for clarification: A node of this custom type X contains a field that points to another node, so whenever a node of type X is created I want a combobox that contains all other nodes.
(Best solution would be to only display the combobox during node creation, and no longer during edit. But I could also live with it if the combobox was shown during the edit as well.)
I have tried to customize options_select by defining my own data type and implementing hook_options_list accordingly. The combobox was displayed during creation with the correct values, however, I could not save it.. I have no idea what went wrong there, but on the first submit it would change to a different theme, and when I tried again I got an internal server error. Am I on the right track at all with defining a completely new data type for the field? there surely must be a simpler way?
You're right in that you don't need a new datatype. Here's a good tutorial on how to do this. It's not specifically for D7 but I didn't see much that wasn't still applicable. There may be a better way to do it in D7 specifically but I would love to know it too if so :)
The tutorial linked by allegroconmolto sent me on the right way. Thanks for that.
Here's the simpler way of doing it: tutorial
Basically, it is, as I assumed, a common problem and hence a simple solution for it was included in the webform module by now. It provides a hook_webform_select_options_info which can be used to register a callback method. The callback method is then called each time a corresponding option select of a webform is shown, so that you can easily fill it with the results of a dbquery or anything else. Works like a charm and takes next to no time to implement.