This applies to LUIS (MS Language Understanding)
Want to handle an utterance in the following format
"I met [a-PersonName] at [a-BusinessEvent] in [a-TimeReference]"
Sample utterances might be
I met Jane Allan at the Product Management Meetup in January
I met James at MS BUild in April 2017
I met Lily Tomlin at Learning UX Meeting in June 2018
The challenge is that [a-BusinessEvent] (the bold bits) will grow over time. Sure there are a couple of recurring things such as MSBuild or Apple WWDC but over time I'll want to have the users extend the list of BusinessEvents available. (imagine having a voice interface that would allow 'add new event called seattle chatbot meetup').
Should this be a list? or something else?
Are there any examples I could learn from?
Thank you
If I am understanding your query right, you are looking to have potentially a free form event names and you would want the ability to extract the event name entity out of the utterance consistently.
If my above understanding is right, you might want to take a look at "Pattern.Any" entity. This gives you an ability to extract data from the utterance where the utterances are well-formatted and where the end of the data may be easily confused with the remaining words of the utterance.
Once pattern.any is established, you'll have to add patterns that use the Pattern.any that you have created in order to improve the accuracy.
Visit this documentation for more information: https://learn.microsoft.com/en-us/azure/cognitive-services/luis/luis-tutorial-pattern-any
For eg., in your case,
I met Jane Allan at the "pattern.any" in January
I met James at "pattern.any" in April 2017
I met Lily Tomlin at "pattern.any" in June 2018
And finally, create a few patterns to improve accuracy.
I met {PersonName} at the {EventName} in {DateTime}
I met {PersonName} at {EventName} in {DateTime}
I met {PersonName} at {EventName} in {DateTime}
Related
I have just going through https://visualstudio.microsoft.com/vs/pricing/ and found the prices to be 45 $/ month or 1199 $/year for Professional subscription. Is this cost per user? if so, do I need to purchase 100 subscriptions (4500 $ / per month!) for 100 developers in my company?
It's per user. It's better described on the following page:
https://visualstudio.microsoft.com/vs/pricing-details/
I'm just looking for the same and you can be glad to have a price tag of $1199 - to me as single user located in Germany this turns into 1540€ on check-out, that would be $1813. VAT is only a small share of that difference.
Btw, the fee is for the first year, renewal will be relatively cheaper.
With your larger number of seats, you should also investigate the SKU 77D-00092 and 77D-00095 respectively. As I see them advertised in various places these still come with two years of MSDN subscription rather than the one year offered on the MS web site, also at a lower price. Maybe they are missing bundled Azure credits? Those SKUs appear to require an "Open License" plan with at least 5 seats - that's as far as I got.
I am trying to setup a system where I can undertake a search on a document so that any questions can be answered. For example, the document may be about visiting a city and include information like key sites and opening times. The user would be able to ask something like "What are the main places I should visit" or "what are the opening times of the X museum". The problem with this is that it is all very manual and any errors have to be picked up by me to manually fix. There is always only one document and it can be anything from 5 to 50 pages long. Does anyone know any AI tools that could help or how I could introduce self learning?
I have achieved this by creating a list of keywords (tags) against key facts. So the fact could be What are the Museum Opening times? and the keywords Museum and Time. I then undertake searches using AND and OR functions against the keywords. I have also created a database of pseudonyms e.g. for time I may have opening hours, times of opening etc.
What are the opening times of the X Museum?
The opening times of the museum are Monday to Friday 8am to 8pm
There is not an awful lot of detail in your question but it sounds like an application for a chat bot such as ALICE. To create a chat bot for your document you could use AIML.
Here is a tutorial for this :-
https://www.devdungeon.com/content/ai-chat-bot-python-aiml
Noting that I personally have never created such a thing it does seem that the functionality that you require is a good match to this technology.
I have included two pre-built entities in my app, a DateTimeV2 and one Number.
I tried creating an intent "QueryByDate" and added this utterance:
Book Me 2 tickets on 12/21/2017. I was expecting 2 and 12/21/2017 should be identified as the prebuilt entities, but they are not.
Please help me in understanding what am I missing here.
Thanks,
Subrat
It's a bit buggy. All these services are in beta so expectations should not be so high :). Go ahead and add two or three utterances. Then go and train the app. When you go back to the intent, the entities should be recognized.
I've interacted with LUIS for many hours and I must say, if you want to do natural language processing use Facebook's WIT.ai. It's an excellent platform!
I want to figure out a way who have talked {view, like and comments) about a specific post on our Facebook's fan page. we want to examine the influence of the daily activities of fans on the timing and amount of a specific post. Influencers and the occurrence time of every "like" or "comment" should be collected. Is there any tool for modeling and clustering of post's data?
Thanks in advance.
This is definitely possible. I know this is not the forum to advertise a specific product, but I literally wrote this exact thing over the last couple months for our company. Check it out http://napkinlabs.com/apps/ it is called SuperFans.
I have released an app which,. within its functionality displays date and time strings.
I am aware of the differing formats across cultures - however in some cases I had hardcoded values- for example I had gone with a custom format that was the 12 hr clock and showed AM/PM
I am now changing to use the standard date time format strings where possible, and so, for my times,I am now using the shortTimePattern.
What has surprised me is that for the US this shows as say 3:15PM but in the UK its 16:15 i.e the default there is the 24 hr clock.
Similarly in the US the long date includes the day of the week, where as in the UK it does not.
I am thinking that these defaults must be right and are what is expected within that country but is this really the case? I had no idea that the UK default would be a 24 hr clock. And, for those users in the UK who have the app, will they be annoyed when the next update shows the time in this format?
Interested in any opinions around this.
thanks
UK Users will not care between 24hr and am/pm (we don't talk to each other saying "It's fourteen hundred o'clock :P).
Dates are also fine unless you're using format of 12/02/12 as in the UK that's considered 12th Feb whereas in the states it's December 2nd.
This is not the place to solicit "opinions" on whether the behaviour provided by the framework is correct.
Assume that the framework is correct, unless you know otherwise.
If you want to know what your users will think you should ask them. (I assume that your users are not typical users of StackOverflow.) If you don't already have beta users in your target markets to ask then make the change to use the "standard" behaviour. If there is a problem your users will tell you.