The Wit AI project added a new concept called Story to their Converse functionality. Is there a way to manage (create/edit/verify) these Wit AI Stories via their HTTP API?
This is what the Wit.ai team said when asked about having an API for stories:
"Not yet, unfortunately. The only workaround is to use the export/import and update the JSON files but I agree this is not great. Or to try to emulate the calls made by Console...very hacky sorry."
You can try API.ai, their API is more complete and it's free as well.
Related
I would like to create my own web chatbot and i like to integrate my app with wit.ai for natural language classification.I need to know how to integrate wit.ai service(through api call) with my application(any language in backend).i am using C# in front end.I have gone through the integration part Which posted in wit.ai website.But i don't know how to connect it .Could anyone send me a integration details little briefly
I think the short answer is its similar to how you would call any other APIs from your application server components. Wit exposes multiple APIs like message, speech and converse which you can call by passing the Authorization token and other payloads and make use of the API response in your application.
You can use message API if you are only interested in extracting
intent and other atributes of the sententense
Use speech for building voice based application and
Converse if you want to build a little more smarter app. Currently you can only pass text for converse APIs.Hoping they will introduce voice option for this soon.
Now to make things simpler, they have also provided SDKs in various languages like node-wit, pywit etc. So if you want to build your server side logic using on nodejs or python you can use these SDKs. The advantage is that you dont have to manage raw APIs calls and instead it is all managed by SDK. Also, other big advantage is that you can make use of runActions method which encapsulates converse API and make things simpler. If you want to build in nodejs then the messenger example is a good starting point. You can borrow all this logic/concept in your app and replace FB related calls etc with your custom bot. For Python you can look at the below link
https://github.com/wit-ai/pywit/pull/55
Also, you can explore the options like using other frameworks like botkit if you plan to integrate wit with other chatbots like FB messenger or slackbot as these frameworks provide more flexibility and ability to easily switch to different chatbots in future. But they don't seem to properly support the converse API of wit.
You are specifically looking for integration details. Since you are using c# for frontend app, natuarally the best option would be to use c# for backend as well. In which case you will be left with directly calling wit APIs from your backend as I think there are no SDKs in c#. If you want to make use of SDK in node or python etc then you will have to build a rest based backend (for example) which can be invoked from your c# application. I am currently working on a nodejs app and integrating it with wit using node-wit. I can share some code once its ready but i dont know when I will be able to finish it. For bootstrapping my application I have used this node application. If you have some understanding of node then you can look at the /server/controllers logic. Similar to this application I have built a witController which uses runAction to interact with wit and I am calling this from front-end when user submits a message to your bot. The biggest challenge in runAction is to figure-out a way to send back the wit response to your front-end and get follow up response from user. Wit sends the response in Send method as you can see in the node-wit's messanger example.
Hope this helps!
How to use BigQuery from elixir code?
I would like to use Big Query web API from elixir code. Although there is no elixir client SDK in SDK libraries page
I will use BigQuery API with service account. For using service account, I have to exec "JWT encode"...
Do you know suitable elixir library for using BigQuery API?
BigQuery implements a traditional REST API. If no one has implemented it before, it should be straightforward to create one using a REST library.
REST for elixir:
https://github.com/h4cc/awesome-elixir#rest-and-api
You'll need to authenticate too, OAUTH2 for elixir:
https://github.com/h4cc/awesome-elixir#authentication
It's not a library, but the following repo has some example code for some basic usages on BigQuery (It was just an experiment and pretty much immature, but I'm hoping that it's better than nothing..)
https://github.com/parroty/big_query
I know this is an old question, but looks like this exists now: https://github.com/googleapis/elixir-google-api/tree/main/clients/big_query/
Writing this for folks who come here years later (like I did).
does the old api of onedrive (skydrive) which was exposed in liveconnect api will be down(close) with time???
the above statement was made in one of the blog
http://venturebeat.com/2015/02/24/microsoft-launches-new-onedrive-api-with-change-tracking-resumable-uploads-and-customizable-thumbnail-images/
m currently working to integrate the onedrive with our product, it is reallu tough decision to go with new apis or use older one...
pls comment, ty in advance
We recommend you use the new API. It offers more features, better performance, and should be easier to use than the old one. We haven't announced plans for the LiveConnect API, but we're actively developing the new API.
I'm working on a cloud project and I have to use Contrail it provides a RESTful API to use, but it is documentation is too small. I would like to know how to perform rest requests, especially post requests on "project" resource.
Each Contrail config node provides supported API details(searchable) at http://:8082/documentation/index.html
You could also check here which is a bit old but should serve your project API needs.
Your question is old, but as there are no other answers I thought I would contribute one. Were you interested in an answer or just wanted to share the documentation you'd written? I agree with you that Contrail documentation is very scarce, and every attempt to contribute to it should be appreciated. I had to integrate with Contrail API a while ago, and found out that in some cases I had to call OpenStack APIs to get the job done (Contrail is built on top of OpenStack). Project and user management were some of them.
I wrote a documentation about Contrail api, For each resource I mention the allowed methods and which parameters to integrate on the request body and I test it until I get it to work finally I put an example. I haven't finished the documentation, I stopped working on Open Contrail. The documentation is my personal try maybe it would be useful and maybe it contains some mistakes, Thus, I would like to share it. You can find it here.
I need some help with implementing a status board with a REST API. I've already found the two best options (in my opinion): Stashboard and Whiskerboard.
I have already tried Whiskerboard and I find it easy to install and deploy because it does not need Google App Engine's SDK to be used locally. However, I need something like Stashboard's REST API: http://readthedocs.org/docs/stashboard/en/latest/restapi.html
I know that Whiskerboard is based on Stashboard so probably they have the same REST API. But I cannot seem to find any documentation on Whiskerboard's REST API. Moreover, I looked into each status board's models and there are lots of differences, with Whiskerboard's being largely simpler compared with that of Stashboard's, so they must not have the same REST API.
I will appreciate it if anyone can provide me a link to Whiskerboard's REST API which should be much like Stashboard's. If there really is none, some other solutions could still be suggested. Thank you in advance!
No, it doesn't have the same REST API. You could find this out by reading the code. If you want a similar REST API you have the following options:
Code the feature yourself.
Ask the developer of Whiskerboard for the feature.
Ask someone else at your organization to code and contribute the feature.
Good luck.
You can check out this solution forked by sijis and myself, jvasallo, that we have been working on. He forked the repo for whiskerboard and added an API to it. Still uses Django, and recently I did some work porting it to 1.5.1. Initially, Whiskerboard is a 1.3 Django app.
https://github.com/sijis/whiskerboard