Hi, my name is Killian Lynch. Today we're going to cover the Data Insights and catalogs tool, so as you know, the RDBMS already has many machine learning algorithms built-in and the analytic view is what allows the RDBMS to understand the user's intended use of the data in the RDBMS. Analysts typically perform what are called slice and dice on top of these analytic models and in this feature, we've essentially tried to automate that slice and dice process. We report findings that are statistically significant. Think of these like anomalies in your data. Now clearly we don't have the business contexts for these anomalies the users do. We're just here to help the users get to these faster. Now let's take a look at the Data Insights tool itself, we start by saying that we want Insights. We'll start with the QTEAM schema and run that against the business models that we just created and use purchases as our measure and press "Search". By scanning through the data and running a large number of queries, a number of candidate Insights appear. For example, here's one for the taste of middle age females for various genre of movies. Now, each of these bar chart is for an individual insight, so we have quite a few, as you can see scrolling through. Here's one that may be interesting, movie genres consumed in June. Drilling down on this, for the month of June, the actual purchases are shown by the blue bar. The expected purchases are shown by the green line and three bars have black borders. These are where there's a large difference between actual and expected values, so the data is actually suggesting that as we move into the summer months, the taste for movies shifts from Sci-Fi to a lighthearted comedy and Romance. We can actually take this Insight and next year's planning campaign, we could anticipate this shift in sentiment with the changing seasons. Data Insights is a remarkable tool. I want to reiterate. The autonomous database uses this knowledge of the business model to automate those slice and dice processes that we were talking about earlier. The tool identifies those anomalies using machine learning algorithms and then catalogs them for the business user. This is a great feature for new users of autonomous database. Imagine creating a data mart and then getting interesting results right away autonomously. The last tool I want to show you is the catalog. The catalog has a wealth of information about the data we've been working with. First, we can see all the tables in the schema in our card view. There's also a grid view, which is a little bit more compact version and there's a list view. You can pick the views over on the right side by simply clicking. This would be grid view and then this is what tables field looks like. I apologize, list view. Now for the purposes of the demo, I'm going to go ahead and use the card view. There's also a browser-like capability, so if I type in movie sales, only the matching items are shown. Currently, we're reviewing entities of type table in our schema, but we can actually change the filter criteria. For example, I might want to view entities of type, business model, and analytics field. Now, besides the two tables, we have the additional two cards that we just created. By clearing the search criteria, we now see all eight cards. Now for one of these objects, let's go ahead and choose devices. We can actually look at various amounts of detail. Here's the data viewer. Here's the definition of a table in SQL. Something I find more interesting is the lineage of this table. By expanding the lineage, we can see the table and its columns. We can see that data load jobs by which it was imported and over here on the right, we can see the source file we started with. You can also hover your mouse over each individual link and see the details of each link in the chain. Now I'd like to show you impact analysis and a good way to think about this would be in terms of lineage just turned on its head. For example, let's view the impact analysis for the table movie sales 2020 Q2. Here we see the tree of entities that are dependent on this table. Here's a mapping of each column in the table to the corresponding analytic view and the attributes dimension. We can actually expand each one of these as well by clicking the three dots and clicking expand. Each of these have its individual Insights shown here. We can get actually a great deal of detail if we would like and you can collapse it again to get that higher-level view that you're looking for. That's a quick tour of the built-in Catalog tool. Again, this has a wealth of information about our dataset, so data is capital and the built-in catalog allows you to maximize its value. Data lineage and impact analysis are now at your fingertips with this integrated tool. To summarize, use the catalog to view objects in the autonomous database. There's a browser-like search capability, where we can search and filter to show entities are particular types. For individual entities, we can see valuable information such as lineage or impact analysis at various levels of detail.
