Hi, I'm Seth Redmore. I'm the Chief Marketing Officer with Lexalytics. Feel free to follow me @sredmore on Twitter. This video is going to show you how to analyze your own data. What you're gonna wanna do is download the Lexalytics tutorial on how to run an analysis and Raw Hotel A, at the very least. Before now, you should have already signed up for free, and you should have installed the 32 or 64 bit Semantria, add-in for Excel. Once you've done all that, you're ready to rock and roll and get processing. So, I've downloaded all this stuff. I've installed it, since I am chief marketing officer, this shouldn't be surprising. And here's what's inside of this particular file. So, you've got an ID, date, city, room number, manager on duty, customer name, and then the text. So the very first thing that you're gonna wanna do is make sure that you are logged in as yourself. So, I am signed in with email@example.com. You'll just log in using your own username and password that you gave us when you signed up. So that shouldn't be too hard to remember and there's ways to get to it in case you forgot it. So I am signed in. The very first analysis I want to do will be a discovery analysis, which will give me a high-level view of things. So, what I'm going to do is I am going to select Sentiment, Entities, Facets, Themes. Summary isn't really interesting. And we'll pick Auto-Categories as well, just in case. We will do Manage Analyses and Reports. What we're gonna do is use the default configuration, or for your case, it will be the English configuration. And so the reason why you want to do that is to see what it does without you telling it anything cuz the next step we're gonna do is to tell it about a few entities. So, Configuration, New. Let's make a discovery. Let's call this coursera-hotel-a-32, just because I like that number. And select your source text by clicking that button, select the first cell, scroll down and say OK. Great, go analyze it. It sent off the documents. It's thinking hard about them. It's gonna receive them very, very shortly. This could take a little more or a little less time depending on the time of day, see? There, it's back, 100 documents are processed. Let's look at what it's gonna tell us. Here's the source text. And let's close this. Hotel facets and attributes, these are subject and object, basically. And what it is is someone's giving me the opinion of hotel and no negative facets about it. Attributes, great, first class, amazing, beautiful, nice. That's great. Now we get down to staff, we do see a few negative facets, rude, that's probably where that's coming from. And then service, turndown, superb, excellent, none of those are negative terms. So we're gonna have to dig a little bit deeper into this in order to figure out what exactly is going on with the negativity with respect to service. Other things you'll see here, you'll see the top themes, front desk, great view, nice hotel, extra money, extra money, it's a little spendy. And something I wanna point out here with entities is that hotel A is not appearing as an entity, so we're going to have to tell it. And the reason for that is we anonymized most of this content so that you would have to go through this process, because sometimes natural language processing systems don't pick up entities. One thing that's funny, it's a mistake is, of course, Eggs Benedict, it's being referred to as a person. It obviously is not a person, it was probably capitalized in the text and so therefore, it's a perfectly reasonable name. Great last name, funny first name. But this is the sort of thing that you will run into but with enough data, it tends to get pushed down to the bottom. So let's go ahead and do a detailed analysis of our content. First, we're going to want to put hotel A in as an entity. So I'm going to switchover here to SWEB, which is online.semantria.com. You log in using your same login as we just used to log in to Excel. What you'll want to do is you'll want to make a new configuration, create it from scratch. You want to give your new configuration a unique name, coursera-32. You're gonna wanna give it the type of content, this is customer support and survey verbatims. It's in English. We're gonna say Semantria for Excel and then create the configuration. I'm not gonna bother because I already have a configuration inside of here, coursera hotel A, that's what we're gonna work with. The first thing you wanna do is go into the little gear box. Make sure stuff is set sanely, particularly in detailed analysis. You want these to all be around like 20. And the reason for that is, if you don't have these set high enough, you're not gonna get anything returned. It doesn't matter if there are exactly 20 or not, it really, really doesn't matter. You just want it to be enough that when the documents are processed, you get enough back per document. So note, I really don't care about how many there are here. Sometimes you're gonna wanna make it more precise than other times, but really, I don't care right now. So we'll save that, great, and then we will come back out here and we will look at the entities. This is equally important. We need to tell the system about hotel A. So, we are going to, if you don't have this, just click add row. We're gonna say it's a query, that's a plus. And the query is looking for quote, Hotel A. That's exactly what you're looking for inside of the content. Normalized value, Hotel A. Entity Type, Company. Label could be things like competitor or hotel, that sort of thing. The normalized value, this could be, the search term could be H-A, or hot a. Think of stock ticker symbols, those are also names for companies. You need to be able to normalize them to a single name, so we're choosing Hotel A as the single name for that. Save this, saved, and Publish. Publish it all. Great, it's now live. Let's go back over to Semantria. Reload the list, just in case. Great, it's all there. We're gonna use coursera hotel A, close this up, now let's do some analysis. New one, DETAILED analysis, coursera-hotel-a-42. Source text, same thing as we did before. Just select this text. There, now, optional ID, this is important if you are actually importing this into something like Tableau or Qlik or any of these other visualization programs, it will let you join things on an ID. So, there we go in next column, all sewed up, detailed, yay. Analyze it. Great, let's open it up. Now we can see hotel A is appearing as an entity, excellent. Let's close this to get this out of the way, all kinds of good themes inside of here, I'm interested in a couple things. So, first and foremost, we can see that hotel A is appearing and that's great. So, we can do analysis with hotel A and actually see how often hotel A is being referred to in a positive, negative, or neutral way. I'm more interested in themes because themes are actually kinda cool. So, what you wanna do is, first, click there. Then, Insert > Pivot Table. New worksheet, OK. We wanna sum up the themes and look at the themes. So we add the themes to it. We're gonna add the themes to the sum. We're going to take Theme Sentiment plus minus and use it for a column. Now we can look and see. It doesn't look to be anything too bad in the negative column, that's the stuff to the left. But if we wanted to actually get a feel of what's happening with the front desk, just double-click on that and we have everything. Source text, etc., etc., etc. So, based on past experience at Vegas, I booked at hotel A because I wanted top-notch service and a worry-free visit. This is not what I got, yadda yadda yadda. Clearly, something negative. So the theme inside of here was front desk and it was negative. So they had a bad experience with the front desk. What's really cool about this now is I can say ID eight. Let's go here to the sentiment, to the overall. Who was on deck for ID number eight. It was James Haddock. So, find out what happened with James. Was it a customer complaining and being a pain? Or was it James's issue? That's what I wanna show you. Take your own data. Do the configuration yourself. You can actually make your own categories. You can, so inside of here, if we go back to, here you can set up your own queries. These are boolean queries. You can set up and, or, not queries. You can also set up change things with Sentiment. So if I wanted to add a word like mold as being negative, I would do that in here. So you have a lot of ability to tune in inside of here. But just the baseline functionality will tell you a lot about what's happening in your hotel. So give this a try. Try it with your own data. And again, if you are looking at something commercial, I highly encourage you to give us a call and contact our sales team. And we'll qualify you and give you support and perhaps some more credits, and work with you on this. So, thanks again, I'm Seth Redmore. And good luck with the rest of the MOOC, I hope you enjoy it.