So that’s the second example of how data visualizations are quick at solving a problem. The concept here that you’ve seen in the demos is how charts that you put on pages address business issues. Combinations of visualizations then show a story of performance for a unit or for a problem. There is a set of charts we work with.
You’ve seen bar charts that have a variety of looks and fields. A pie chart shows relative performance. Line chart show trends over time very well. The map shows distribution across a geographic area. It could also be a logical map. It can show links between things. So there’s a set of things that maps do really well.
The heat map you just saw is another type. We have seen visualizations for text data. The text data would be selectable. There are different ways to cut the data down. A scatter plot you saw with the mutual funds but that can be turned into a bubble chart so you can resize and shape it for different kinds of data. Time table show transactions and in this case you actually see for this specific element. There are a set of clusters of transactions and then they stop and they pick up another areas over here.
| #1 Ranking: Read how InetSoft was rated #1 for user adoption in G2's user survey-based index |
|
Read More |
Another statistical view shows you metrics really well with means and standard deviations. A multiscape is a 3D view histogram. A data constellation is a view that shows affinities and clusters. You might use it in social networking to see who’s related to whom. It also can be used in a network topology that may be used to see how network traffic flows and so forth.
But the point of that was there are a set of charts that are able to help you see your data in a way that helps the data tell the story. Quickly a third example is with an airline network with overbooked flights. The data consists of single flights. You wouldn’t just want to sort a list to see the worst flights. What you care about is how groups of them do.
So here’s an example of a visualization that takes a major carrier’s North American flight Network, and the pillars that are on the airports, the bigger the pillar the more over sold are its flights, and the darker the color the more each flight is oversold. So you look at this and something jump out with a lot color and with the higher columns is the northeast. This is where there are more overbooked flights.
 |
Read what InetSoft customers and partners have said about their selection of Style Scope for their solution for dashboard reporting. |
Some of the flights segments, like Boston to Miami is a problem. And Miami to LA is a problem. So it’s not an individual flight. It’s the pattern. So in this case, you can clearly see that Boston’s having a problem. You can click on Boston, and you can see which specifics flights in Boston having the most problems. And you can go the detail page and actually export the list of flights which you want to look at and take action.
Maybe the problem is in that condition, if a storm comes through, you can have a problem with a lot stranded passengers. That’s an example of where you can picture how a long list of delayed flights would be really hard for people to pay attention to and focus on a pattern. And in this case seeing the pattern of the northeast, you might make different decisions after seeing the worst flight which actually were some of those going to Miami to Los Angeles and so forth, you might change what you do by seeing the data in a way that tells a story.
With the couple of minutes we have left we’re going to quickly look at predictive analytics, which complements the other visualization technology. Here it’s using mathematical tools to determine patterns in one set of data in order to predict behavior in another set of data. It integrates really well with data visualization. Let’s take a look.
So here the process is you select the target to compare a base population. You look at explanatory fields which can be in the main table, or because we have all the data in-memory you can calculate them in the other tables and put them in the main table. So there may be 20 or 30 core tables of contacts, and the main one is the people where you’ve got information on degrees and employment and activities, you can create factors on those tables and bring them back to the main table and use them in a regression.