Case Study: Indiana University, Data Virtualization, and The Initiative for Decision Support

“We can use business intelligence tools to solve problems, and we can have really smart technicians write code, but if it doesn’t serve the needs of the business and actually answer questions that people need to know, then our work is pointless.” Dan Young, Chief Data Architect at Indiana University, said this (IU). Young said that IU was able to answer these questions for a lot of different schools and departments by using data virtualization from Denodo. (data science in Malaysia)

It’s call the Decision Support Initiative (DSI), and Young said that it’s a “opportunity for IU to change how we use analytics and the way we make decisions.” Indiana University, which has eight campuses in Indiana, has more than 19,000 employees and serves more than 114,000 students.

Assessment of data (data science in Malaysia)

IU has had a hard time finding accurate and timely data for making decisions in the past. There were “multiple copies of the same data” in their data warehouse, he said. It had been there for 15 years. People in different departments and campuses had different definitions of the same thing, which made it hard to make important decisions that needed timely, relevant, and accurate information. Many important university decisions are made by people who have to deal with a lot of different kinds of data, in a lot of different formats, and in a lot of different places. They might not even know where to find all of this data, or even that it even exists,” says the teacher.

Young wanted a project that could be done in an Agile way, but also wanted to figure out how to make big decisions without having to guess. “We really wanted to try to focus our business intelligence and data development work on the idea of Agile BI,” said the people who worked on the project. We wanted to try to keep giving the university value over and over again.

A key part of the solution he came up with was the idea of incremental delivery: “Taking what you have and making just a few slices, or parts of the whole picture, so that you can show value and people can start using it as soon as possible,” he said.

Traditional data warehouse (data science in Malaysia)

When you do a traditional data warehouse project, you would spend months gathering requirements and writing all your documentation, and then you would build the data and then you would build the visualisations. Is your product good when you show it back to the person who will use it? It’s what I asked for, but I no longer need it because my needs have changed and it’s been 18 months.

To help the University make better decisions, the Decision Support Initiative was born. Its goal is to provide timely and relevant data to help people make better decisions.

Development of a programme

Young looked at a lot of tool-object relational modellers, like Hibernate and.NET/Link. But when he thought about recording tools, he came up with some new ideas.

One of my challenges was to look at the technology space and see if there were tools that could help us become an Agile BI type of company.

He kept looking, and he found a technology called “data virtualization.” There are different tools for data virtualization than there are for data aggregation. They were made to help with this “Agile” way of putting data out there.

It was while they looked at technology and tried to figure out what might work that they started to build a team and hire developers. Denodo was the best choice for us as we tried to move forward in the Agile BI method in June 2015. We looked at all the options and decided that Denodo was the best choice for us.

CMO Ravi Shankar says that his company has been working on data virtualization technology for 20 years:

Virtualization has been around for a while now. A lot of people used to call it “Enterprise Information Integration.” Now they call it “data virtualization.” So technology as a concept has been around for a long time. It’s just that the names have changed as the technology has changed.

Data virtualization brings together data from different sources, places, and formats, but doesn’t make copies of the data. This creates a single “virtual” data layer that can be use by multiple applications and users.

Uses Data Virtualization: How IU does this

The Decision Support Initiative allows people who make decisions at the University to ask for a “charter” or project, Young said. A person requests a report or dataset online by filling out a form that asks about “the business problem it will help us solve.” The form asks for this information.

If this charter is choose, “he said: “If this charter is choose, here are the business people that we are willing to work with and help you define requirements and work through that process with you.” This means that the project will have a person who will work with the business analysts to make sure that the requirements are turn into “incremental deliverables.” The charters are then look at and prioritise by a steering committee.

One example of what the DSI can do is call “Academic Metrics 360.” This is an example of what can be do (AM360). In what he called IU’s “crown jewel,” the project uses data virtualization to give a “360-degree view” of an academic centre, which is what the project is called. They can look at how many credit hours they are teaching, what they are teaching, how many students are taking these things, and how that all adds up to money and funding. The goal is to make sure that “all divisions can essentially have the same information,” so that each can take that data to annual reviews with the provost to show why they should get more money or show how they’ve done well or had problems, he said.

Source: data science course malaysia , data science in malaysia