 Thanks, Martha. I'm happy to talk a little bit about how we use Tableau at UMass, show some examples from a couple of different data sets, and then give a little demo about showing how we actually use Tableau in the software itself. There'll be time for questions and some other demonstrations within Tableau if there's interest and we have time. So, in terms of how we use Tableau at UMass, we use it in three different ways. The answer question wins the best time to recruit for focus groups or informed decisions. When should we add an extra student to a service desk? To monitor data over time, in a monthly report fashion, also related to service desks or different collection related projects. And also to communicate results, sometimes as with a PDF or sharing examples of what was done with different campus stakeholders. And for me, using Tableau is sort of my preferred default. If I have a new data set to work with or something to look at, I've really become accustomed to trying to interact with it from the beginning in Tableau. So here's an example of our live analytics service desk statistics. So we use live analytics to collect data at each of our service points. We download that monthly into Excel and then build a variety of visualizations with Tableau. So this is an example of activity at the Learning Commons desk. It shows by day of week and by hour. It shows comparison from FY 13, 14, and 15. And the filters, the little radio buttons on the right, allow for selection of different desks so that that automatically filters and rebuilds the view accordingly. We have a lot of other views that we look at as well that show the differences between students answering questions, the scale, the level of difficulty of questions. But this is an example built off of Excel data that's updated monthly. So this next view here is of some data that we have from our Learning Commons. So we do headcounts three times a semester every hour for a week. We go around and we count users in different locations in the library. And this is sort of another way, another kind of view that we build. So here the dark green shows the highest percentage. So we were looking at when our group study rooms, how often is there nobody in the room? How often is the group size of one or two? And how often is the group size of three or more? So thinking about the size of the groups in each room and how frequently we see them used in that way. So this is just another view. And this is also Excel data. Here's a little view of the question I mentioned earlier. We had some outside folks coming in to do some recruiting for a focus group. And they said they could come on Tuesday, Wednesday, or Thursday. And they wanted to know what time to come. So we very quickly pulled this up, sent it off in an email, and they scheduled their recruiters accordingly. And the nice thing about this is that if we hadn't already established a connection in Tableau and hadn't already had our gate count data easily available, it would have been really cumbersome to try and answer this question. And in fact we might not have, because it just would have been too complicated. But instead this took only a few minutes, and we were able to send it right off. And it was really helpful for making a practical decision. And this is also an example, this was a Microsoft SQL server, and it was a live database connection. So we can look at the gate counts every 15 minutes, I think, approximately. This is an example of a connection we make to our ILS's ALIF. It's an Oracle database connection. And this is a visualization of some collections data. So in this case on the right hand side you see where you can choose an order group. So the selector for math and physics would choose her order group. And then she'd choose the budget years that she wants to have shown. And the monographs, the books that she selected are visualized here for those order years. And she can tell how much money she spent for items that circulated and items that didn't circulate, as well as the number of items purchased and the percentage how that breaks down in terms of circulation. So that's the view at the top. The selectors are also interested in the extent to which items that are purchased are duplicated within our five college consortium. So we have an effort to minimize either unnecessary or unintentional duplication. So you can see here whether the items are unique or if they were also purchased elsewhere in the consortium. So this is a particular view, but you can get two other views that go to title level detail and provide additional information. But this allows the selectors to choose their order group and to choose budget years and keep track of the monograph use that way. This is an example of a dashboard that we have for EBL data. This is also Excel data. So we get that updated monthly. And we have a five college consortial pilot project to use EBL eBooks. And we're interested in, this is sort of an overview dashboard in terms of what we're spending on short-term loans versus auto purchases by user status, by publication year of the item, and by subject. This is nice because it can be filtered by each school so we don't have to duplicate the analysis at each institution. We can look at the consortial view or at the individual view. There are also other visualizations that show publisher information, title information, different kinds of use breakdown. So we've been looking very closely. We adjusted the cap of our price for what we would pay and we'll be able to track monthly how our spending has changed as a result of that price adjustment in response to the short-term loan increases from publishers. So this has been really practical too. And it's pretty powerful. I've been thinking as libraries are doing more cooperative things to have views like that filtered and sorted in different ways for different parts of the partnerships is really great. Does it save time? This is an example of how we've visualized some of our minds data. Many folks are familiar with that, measuring the impact of network electronic services. So this is a survey of e-resources, asking users about the purpose of use, where they're located, their school and college affiliation, why they selected the resource, what the resource was. And we get this data through a live MySQL database connection. So as I think about the range of connection sources that I've shown is one of the reasons that I like Tableau because we don't have just a single platform or a single data structure. Our data is varied. And Tableau lets us connect to it in a common way, even though the source or even the location comes from different places. So in this case, the filters on the right-hand side, we can say we're working with someone from nursing and we want to see how those resources are used in support of that department, and we can filter it that way. Or if we want to look at what undergraduates are doing, we can filter by undergraduates. You can really do sort of the equivalent of multiple crosstabs. We could take a look at what people are doing in the library versus out of the library. We could look at how sponsored research is being used compared to undergraduate coursework. And we also have the time and day of week information as well, which is interesting to put together with our gate counters. So we see when people are doing things electronically from off-campus, and what kind of traffic do we have in our building as well. So in terms of how we share our results, we published some things to the Tableau Public Server, and that's available to the world. We also have our own version of Tableau Server, where we have 10 user licenses, and we have some data that we keep there, some visualizations that we keep there. We will sometimes print views, either as PDFs or as images, that we'll combine into other documents. I email images to people if we're asking a question and just sending an answer. Here's a snapshot of something. And quite frequently, one of the advantages of Tableau is there is this sort of online interactive capacity, which really is great, and that's something that we saw maybe in the collection development selector dashboard. But oftentimes, I'm working collaboratively with somebody at my desk, and I'm using the desktop professional, the software to develop the views, to answer questions. We're looking at data together. We're thinking about, you know, how does it work? What can we know? So it's often we're doing it as part of the process rather than just publishing a result. And so we use it as a tool to interact in that way. So what I think I'll do now is I will show a little example of building one of the dashboards. In fact, it'll be, we'll build these bottom two Time of Day and Day views so that you can see what it looks like. There is a question while you're pulling that from Melissa Horowitz at MIT. She's asking how do you get from the Excel data into Tableau whether it's coming from one database into Excel and then into Tableau, or whether Tableau reads directly the Excel file? Tableau can open directly Excel files. So if you have an Excel file on your desktop, you can say the equivalent of file, open file type Excel and it will read it. So are you seeing that okay here? Shall I run this? So this is a little example of building by year. So there's a lot you can do with year, date, and time and how I'm showing this isn't necessarily how I would build it myself but you can see a few extra steps. So here I'm choosing our second weekday. I'm looking for weekday here. And then you can toss those off of the row. And now we have weekday. And I'm taking the number of records and it's going to give a number. And I almost always show the label marks right away and I almost change it always to bar. So people will see by now and find the horizontal bars. And there's the view. So then you would take your sheet, you would rename it. So many of the right-click features and the way you move around will be familiar to people. I'm going to make a new visualization. This time I'm using my date and I'm going to change it to hour. So this is a 24-hour period. We can go all the way around and viewing it there. I'm going to show a couple. This is where you can try different things. You can try that. You can show your marks. I'm going to flip the axis. And I'm going to reformat those numbers because I like the 12-hour AMPM. It really helps me look at the data so I can see that. So that was really easy to do too. So I'll rename that. And now I'm going to go down to a dashboard that I already started, already put together. And I'm going to add these two views to the dashboard. So you get a dashboard screen and you drag any of the views you have onto the dashboard. And you have that there. So they're there. And you can do some adjusting for size or a little bit of formatting. And you can change what you show. That's more of a normalized view. So that was, it was pretty quick, you know, in terms of some of what you can do, how to build that. Another, I have another little demo here to show you which was an attempt to show what I'm more often doing than, you know, building a nice smooth date view like that. It's really more of an interaction with the data and say, huh, what will this data tell me? What views should I be building? What questions does it help me ask? And this is going to look at the minds data also and it's going to sort of do the user group by school and college. I want to think about what do we know about user group and school and college. And I'm going to put them on the rows in the columns and I'm going to put the data in. And then I'm going to explore. These are the options for the different views. That doesn't tell me much. And then 744 undergraduates in natural science. That was sort of the heaviest concentration. I go back to my favorite. I try something else. It automatically gives me some color. That's a lot of color. I'm not sure I like that. Let me try it the other way. So now it's looking a little more interesting like something I can understand. I'm going to put in my label marks. I'm going to try and move things around again. Try it the other way. I don't really like that. Now that's more interesting to me. I'm starting to think this is interesting. I can understand the color a little bit better. And I can do this kind of click filtering really easily. But I'm going to, or click sorting. But I don't want to see the other. It's distracting me. It's too many columns. So there are many filter opportunities. So you can show your filter and then you can get rid of things. So I don't want to see the other. I want to see my three groups. Now I can follow along a little bit. That happens if I sort this way. So this is just a little example of how I might explore the data, depending on what the question is. And the question we were looking was about the use of electronic resources by different user groups. The faculty, the graduate students, and the undergraduates. Right. And the disciplines on the horizontal line. Right. And so this, to answer Lisa's question, if I were to connect to an Excel source, I would be doing something, I would do new, if it were a new workbook. I would choose connect to data. And I would choose Microsoft Excel. And it would bring up, you know, I would choose the Excel file and automatically connect. So that's that. And these are the other data sources. You can use an access database. You can use a text file. And then a whole variety of database formats. So we don't regularly use that we've also connected to Iliad. We've connected to Google Analytics. There are any number of other sort of database connections that can be made directly. But you also saw my examples. I use a lot of Excel files as well. So one thing I can show another sort of aspect is go to location. So we see here, this is sort of the screen where you would build from. And we can see this is the breakdown for off-campus, on-campus, but not in the library, in each of our library branches. And I say, well, those numbers are great, but I would rather look at this from a percentage standpoint. And that's pretty easy to do by adding a quick table calculation, percent of total, and there's my percentage. And I can format that to a number of decimal points that I want or don't want. So there are a number of quick calculations that are already built in that make it really easy to change the way you're thinking about or checking your data. You can also do customized calculations, which can be a little bit more involved, but these are pretty easy to do right off the bat. Another thing I thought I would show is your ability to change the filter. So here I often change to a single value list rather than a multiple value list. In this, this is faculty and graduate students, but you might not know that if you don't have the affiliation listed on your view in some way. So this is all. And I change it to a radio button single select. And I can see here that 75% of graduate students are off campus. 70% of faculty, but only 50% of undergraduates. And that sort of radio button single select makes it a lot easier to filter. Great. There is an extension to the earlier question. Lisa has a follow-up. Whether you can pool in the Excel data together with, let's say, data from ILLIAT data. So in other words, if she wanted to see the consortium borrowing with the local ILL borrowing in the same dashboard, would that pooling together happen in Tableau, or do you have to prep the files outside Tableau? Well, so there may be sort of, maybe two answers to that if I'm understanding it correctly. You can sometimes blend or join data if it matches up nicely so that you can take two different data sets that share a field or a key that you can somehow connect. You can join that data. There's the ability to do that. And then you're working with it within, it becomes as if it were a single data set. Sometimes that's a little more complicated or it doesn't work quite as nicely as you might intend because of issues with our data. So what you can do is you can connect to multiple data sources. So I can connect here to another data source. I'm not sure if I can do it at the moment. Yeah, probably, I won't distract us by trying to do that right now while I talk, but so that you have multiple data sources listed here. So this is the mines data source. I could have also my live analytics data source, or I could also connect at the same time to our gate counter. And then each view, I could change data sources by clicking and build a view. And when I go to build a dashboard, I could take any view. So I could take this and I could take another view that was built off of another data source. So they might not be blended, but you could put them in the same dashboard. Excellent. And Rachel, I know in your case you don't have a university-wide tableau license, like, you know, Sarah and Jeremy from Ohio State and British Columbia did, but you ended up getting your own license. Can you tell us a little bit about the pricing? I'm not sure off the top of my head. I'm a little reluctant to say. I think it might have been around 7,000. And a ballpark figure. Yeah, I think that's the case. They do have educational licensing. And we got that a little while ago, so I don't know if that's changed. We could certainly have more of that information for our follow-up question call. And I'm sure they'd give you that information directly from Tableau as well. Our server allows 10 user licenses. The lowest level that you could purchase. I can say that we have added an additional desktop professional for the building tool for one of my colleagues. Jessica Atomic is now also developing Tableau views as part of her work here in the library. So I'm not the only one who's using this tool for our work. And building expertise is something we've heard from the others too. A couple more questions by Karen Sigone. She's asking if your Oracle connection is live. And you can tell us a little bit more about how Tableau server works. I don't know if that's more technical than you want to address. Well, I can try the first one in terms of about it being live. So one of our ILS database folks builds a table for me out of the ILS data. And it's set to refresh. I'm trying to think if that refreshes automatically every week, I think. Although we could set it to refresh at whatever period was necessary. So that data is updated weekly in terms of the tables that we've built there. When I connect to the gate counter and to the mines data, those are live connections updated within minutes or I think 15 minutes for the gate counter. And the mines data is probably within seconds. When we first started that, I refreshed it all the time. I was watching surveys come in as they came. So that's a little bit about the database connections. And as far as the server, our systems folks administer that for us. And you know basically instead of publishing, so here I can show you the server. You would sign in. Save to the web for publish, for public. Or I can publish it and sign in to our server. And then I sign in and then I save it there. And if somebody wanted to view it, I can set permissions about who can view it and whether they can be an interactor or a viewer. And they would sign in through a web address. They would just go to Tableau.Librate.umass.edu and they would log in. And then they could interact with it there. That's wonderful. Shall we go back to the slide deck or do you want to show us anything more here? You know, all I had left were questions. There you go. Wonderful. I mean what we have seen I think from the three of you is eye-opening how easily one can explore the data visually. As I keep telling my colleagues here, the only limitation is time. It's lots of data in this universe and lots of data in libraries to be explored with Tableau and be visualized in many different and exciting ways. So I want to invite everybody to share their visualizations whether they are done by Tableau with Tableau or some other software. Share them, send them to ARLSS at ARL.org. We will be posting information about the YouTube availability of these webcasts. I would like to thank Rachel. Thank you, Rachel, very much. And thank all of you for watching this webcast and look forward to the discussion in April. Thank you, everybody. Thank you, Martha.