Cognitive Services Episode 4 - QnA Maker
Cognitive Services Episode 5 - Natural search with the Linguistics API
Description
So far in this course, we saw the high level AI concepts and we built a chatbot bound to a LUIS app. Our bot is already able to respond to IT-related questions and can tackle day-to-day conversations thanks to QnA Maker.
While QnA Maker is a great tool for textual questions/answers scenarios, it cannot be used when dealing with documents or data stored within a database and customers are not always willing to relocate their information into QnA Maker even when feasible.
In this episode, we'll see how to take advantage of another Cognitive Service, namely the Linguistic Analysis API to perform Natural Search Queries against external sources of information. I will show two use-cases: querying a SharePoint document center and interacting with a SQL database but the principles depicted here may be used to query any kind of information system.
Share
Download
Download this episode
- MP3 (14.1 MB)
- Low Quality MP4 (14.7 MB)
- High Quality MP4 (32.0 MB)
- Mid Quality MP4 (26.0 MB)
Download captions
More episodes in this series
Modernizing .NET Apps with Docker, for IT Pros. Part 1.
Related episodes
Cognitive Services Episode 3 - Deep dive into LUIS and Chatbots
Cognitive Services Episode 6 - Document routing & tagging with the Text Analytics & the…
Cognitive Services Episode 2 – Creating & Consuming a Cognitive Service. Getting started…
Bots 101 - Scenarios for bots
A Practical Guide to Neural Machine Translation
Meet the presenter bot, a bot that presents about bots
Microsoft Cognitive Services and Xamarin Forms
Overview of Windows Machine Learning
Windows Machine Learning: Hello World (MNIST Edition)
Windows Machine Learning: Models and Features
The Discussion
-
HarshBot This is great Stephane. Good thought process behind using linguistic service. Very useful resource.
-
stephaneey @HarshBot:thanks :)
-
Dimple Kanwar Hi Stephan,
Thank you for sharing the great knowledge.
I wanted to check that what are the requirements to adopt/implement this technology (NLP, POS, LUIS)in a search engine to improve the efficiency.
Thanks -
stephaneey Hello,
I'd say that for the time being, you should have some risk appetite as most of these are still in preview but other than that, you also need to pay attention to the target languages as capabilities (precision/recall etc.) vary from one language to another.
Best Regards
-
Himanshu Hey Stephane, From where we can download the lemmasharp dlls?
-
stephaneey @Himanshu:Hello, you can install the Sparc.TagCloud NuGet package as it ships with the lemma libraries.