How to Run Search Queries with IBM Watson Discovery - Part 4 - Stock News Crawler

Key Takeaways

what's happening guys welcome to part 4 in the series on how to build a stop news crawler using IBM Watson discovery in the last video we went through the data schema that we created as part of our web crawl in this video what we're gonna do is start taking a look at how to run some queries now the cool thing about Watson discovery is that because it's built such a powerful data schema on the documents that it's already crawled querying it is actually relatively simple because all the entities have already been classified correctly or reasonably well so all that really takes is just to write a query and go I want a certain concept or category or entity and it's gonna come back in your result so without further ado let's get right into it and start building some queries alright so where we're gonna start off is in our Watson discovery dashboard so we're all our stock news web crawls or we're all our data source collections are going to be displayed so what we're going to do is hit stop news web crawl so remember in the last video we renamed our collection so we can jump back into here and start taking a look at our data so once that opens up so you can say it's still the same now to get to the web query or to the queries in general there's a few different ways so you've got the query person predefined queries over here which you can just start clicking run or you can also hit queries over here which will take you to the same sort of page now there's a bunch of different stuff that you can actually do on this page so what we're going to take a look at is just basic query in the next video after this we're going to take a look at how to build some more advanced queries so there's a few different ways to actually query or do really basic queries but it all sort of falls in this first box here so there's three sections within of to any query really with what's in discovery and that's the basic query which is this little box that you can see here up so if we hit the bin it's actually gonna take it away each one of these represents one of the three components of a full-blown query that we could run against what's in discovery so this first bit funnily enough is called just query or basic query the second bit is called the aggregations or building aggregate and finally the last bit is called the filtering so let me sort of quickly explain what each of the what the difference between each of those is so the first bit is querying so it's basically a regular sort of search engine type query the second bit allows you to create summary statistics for each of the documents that you've created so say you wanted a top count of the top categories or the top entities within your document you do it within this sort of little banner over here so the last one that you've got is filtering so filtering is very similar to your basic queries however filtering performs a lot faster than querying because it doesn't pay attention to ranking the documents in terms of relevance or importance so what you want to do a lot of the time is add a filter to cut down your document list and then still add a query to sort of get the relevance that you want so they're both still quite important however you're probably going to use queries more often than having a filter but you should use both because they are they're eventually going to enhance the performance what we're going to take a look at in this video is just the basic query so what we'll do because we've hit the bin and removed it we're gonna add that component back so we can hit search for documents and that will bring up this little proper box or prompt box now there's a few different ways to build up your query so you can use natural language so in this case if we hit natural language and say we wanted to find some results for Amazon for example we could type in Amazon in investing and you should get a JSON response here which is great so I would