Solving mysteries using predictive analytics

In-depth Strata community profile on Kira Radinsky

string(0) ""

Kira Radinsky started coding at the age of four, when her mother and aunt encouraged her to excel at one of her favorite computer games by writing a few simple lines of code. Since then, she’s been a firecracker in the field of predictive analytics, building algorithms to improve business interactions, and create a data-driven economy, and in the past, building systems to detect outbreaks of disease and social unrest around the world. She also gave a predictive analytics talk at the last Strata.

I had a conversation with Kira last month about her entry into the field and her most exciting moments thus far.

When did you first become interested in science?

KiraRadinsky_fullqual

Photo provided courtesy of Kira Radinsky

Kira Radinsky: When I was four or five, my mom bought me a computer game. In order to go to the next level, you had solve simple math problems, which became increasingly harder with time. At one point I couldn’t solve one of the problems. Then I asked my aunt for help because she was a software engineer. She showed me how to write some very simple code in order to proceed to the next level in the game. This was my first time to actually code something.

In the army, I was a software engineer. I built big systems. I felt that I was contributing to my country and it was amazing for me. When I finished my service, I was accepted to the excellence program at the Technion [Israel Institute of Technology] because I had already started studying there when I was 15. I just continued on to a graduate degree.

I knew I wanted to do something in the field of artificial intelligence, because I really wanted to pursue the idea of using computers to make a global impact. I was really into that. I realized that the vast data amounts that we produce could be used to solve important problems.

In 2011, thousands of birds fell out of the sky on New Years Eve. People were writing “we don’t know what’s going on”. It was a conundrum. A few days later, a hundred thousand fish washed up dead on the shore. Many people were saying that it was the end of the world because it was the end of the Mayan calendar!

I like to be a Sherlock Holmes with predictive analytics. I saw a mystery and I wanted to solve it. I asked myself “what do all these things have in common?”. On Google Trends, I was looking at the places where people tend to search for the words: “bird death” and “fish death”. Then, I noticed that usually happens in places where people search for the word “oil spill,”…even half a year beforehand. I don’t know if this is the cause, but for me, it was just the beginning of thinking of how we can use different data points to actually tell a story.

Is that when you brought in the New York Times archive?

Kira Radinsky: Well, the challenge with using Google queries and just looking at what people search for is that it only reflects what people are interested in. I decided that if I want to model causality, I have to know what people think causes these events. I obtained all the New York Times articles since the 1800’s, applied a deep semantic analysis on them, overlaid more knowledge from Wikipedia, looking for patterns, and then building a causality graph.

The result was a vast causality graph, which was a key element in the next generation of event prediction algorithm.

For example, I entered the phrase “iPad prices” into the new system, and the output was a free text the system generated:” “iPad prices are going to increase in the next few months”. The system also explained why it found this prediction. The first event to trigger the sequence of events to lead to the increase of iPad prices was a tsunami in Japan. It turns out, some of the factories in Japan are just next to the shore. Then, the system observed a text news discussion that those factories supply key components to another factory in China. This factory in China is the main manufacture of iPads. The system learn that in the past, if there is a shortage of some supply, its prices will go up. From this inference the system inferred that in this case the iPad prices might go up as well.”

Around this time I had the privilege to join Microsoft Research and work with Dr. Eric Horvitz and Susan Dumais. I felt vey lucky as I think they are among the most visionary researchers in the field of information retrieval and AI. As part of my work, I had many conversations with Eric about medical applications of AI. He holds a MD in addition to a PhD in Computer Science. Together we thought about ways of using predictive systems in medical domain and try to find interesting phenomena. To achieve this, we decided to add a notion of correlation in addition to the causality graph. We entered the text “cholera” and were very surprised with the high precision the system achieved with a pattern based on historical archives. For example, we were really surprised that it successfully predicted the cholera outbreak in Cuba, which was the first one in 130 years.

It sounds like you have passion for predictive analytics.

Kira Radinsky: I always had this passion. The next venture is going to be figuring out how I can use this predictive technology to support decision making. There are so many doctors, politicians and decision makers, that affect so many aspects of our lives but don’t have the right data to make knowledgeable decisions. This is why I co-founded SalesPredict with one of my managers from Microsoft.

Today, my greatest passion is about utilizing data to change the global economy. So many business interactions are based on intuition. It is a very interesting art. However, there is so much data accumulated over the years that each business can leverage. I envision a data-driven economy where each company, big or small, has automated decision-making tools that guide them in their business.

In my opinion, a healthy economy consisting of healthy businesses is the basis for a healthy community. Once this is accomplished, a community focuses on the public common good. In order to achieve advances, we first have to optimize the basics. This is why we are focusing mainly on small and medium sized businesses. For me, this is the beginning of a great adventure with very tangible results, which can affect the lives of all of us.

One step at a time.

Editor’s note: this interview has been edited.

Related Resources

tags: ,

Get the O’Reilly Data Newsletter

Stay informed. Receive weekly insight from industry insiders.