Navigating Predictive Analytics: Benefits and Challenges of Using Graph Theoretic Methods Leveraging Contextual Network Information for Better Accuracy

Nick Lim is the CEO and Founder of Sonamine, LLC. Over the past 15 years, Nick has worked in analytic environments with large amounts of data. Leveraging his training at Harvard, he has led product and strategy teams at MicroStrategy, Enpocket and Nokia to build systems that predicted ad clicks and consumer behavior using terabytes of information.

Predictive Analytics: Benefits and Challenges of Using Graph Theoretic Methods Leveraging Contextual Network Information for Better Accuracy

The practice of predictive analytics has come a long way since the advent of operations research. Current predictive models routinely use more than 100 variables, most of them are characteristics of the object being analyzed, including age and gender for customer analysis, or color and text for advertisements. In this article, we will explore how graph methods can be used in predictive analytics.

Graph methods refer to techniques that analyze a network of objects, rather than pie charts and trend lines. Networks are made up of nodes and connections among the nodes. Other synonyms include vertices, edges, and arcs. Graph methods analyze the link structure, rate of diffusion and the clustering of nodes in a network. Many concepts in the popular press such as Google’s page rank, social network analysis, influencer marketing and driving directions all utilize graph theoretic methods.
Real World Use Cases
Why are these graph methods useful for predictive analytics? One reason is that additional contextual predictors can be derived using graph methods. Let’s start with a well known example: Google uses page rank as one of many predictors for the relevance of a web page. The link structure in the world-wide-web network provides valuable contextual information about which pages are deemed most relevant by the web page creators—this contextual link structure is then used to predict relevance for a user’s query.

Another example: If you are trying to predict the revenue contribution of each employee, in addition to the educational level, job responsibility and experiences, you might consider how well the employee relates to the customers or other managers. The contextual information added here is not something inherent about the employee. IBM conducted such a study with 2600 of their consultants, and created a model with revenue as the outcome target variable and three sets of predictor variables.1 In addition to the standard employee and job-related characteristics, two sets of contextual variables were added. One was related to the structure of the web of relationships, and the other was related to how each employee was related to managers and customers. It turned out that graph-derived contextual predictors were highly correlated with revenue.
Enriching Context Where It Seemingly Did Not Exist
One common question raised is how much of the contextual information is already captured in the demographic data. After all, there is a common understanding that birds of the same feather flock together. To tackle this question, a group of researchers at MIT studied how the Yahoo instant messenger network affected the adoption of the Yahoo Go product.2 They discovered that only 50% of the contextual network effect could be explained by the common underlying demographic variables.

This finding has big implications for predictive analytics in data poor environments. A data poor environment would benefit greatly from using contextual predictors that you already have. For example, if you are trying to predict churn in a prepaid mobile customer base where there is little to no demographic information, using the social network contextual information can greatly improve the prediction accuracy and lower the false positive rate.

Another type of contextual network information that can be included into predictive analytics is the underlying form of the social community that one belongs to. In the study with Yahoo data above that was trying to predict product adoption, they found that the chances of you adopting a product rises if more people in your social community have already adopted the product.

Besides social contextual network data that can be added to predictive functions, you can also deduce contextual information. For example, in trying to predict fraud from a set of credit applications, we can construct a network of information attributes. Using the basic idea of that people with bad credit tended to relate to other similar people, some researchers generated the graph-based hub and authority variables from such a network and added them to a support vector machine prediction function.3 The graph-derived context variables improved the fraud prediction by approximately 20%.

Clearly, the benefits of adding graph-based predictors are better lifts and lower rates of false positives. The improvements range from 20% to 100% depending on the richness of data available today. These benefits translate directly into bottom line results to the extent that your company is already putting predictive analytics into daily operations. So what are the practical aspects of putting this into operation?
How Do You Get Started?
Let’s start with the data: Where do you obtain the data to construct such contextual networks? The IBM study looked at the anonymized email traffic, address books and buddy lists. If you were a retailer like Amazon, you could use wish lists, product gift transactions and share-the-love data to construct the social context for predictions. Financial services companies can use the trading data to construct the contextual network among the traders or credit application data to construct an information network. Telecommunications companies can leverage the call detail records. In fact, the explosion of data collection means that most organizations already have this contextual network information just waiting to be accessed and utilized.

One unique aspect of contextual network predictors is that they are usually specific to the target outcome you want to predict. This is the definition of contextual after all. So if you want to predict churn, then you need to understand how the contextual network predisposes a subscriber to churn. Such contextual churn insight will certainly be different if you were trying to predict employee revenue or product adoption.

Since the contextual environment is changing, the window that an enterprise has to leverage the additional predictive power of contextual networks is also dynamic. Most customers will switch vendors very soon after they learn about a bad experience from a friend, so vendors must respond in a timely manner. Practically, this means that the calculation of graph-based predictors must occur more frequently.

Some contextual network information is relatively stable over time. For example, if someone has a preferred takeout restaurant, she will likely continue to call that restaurant even after she gets a new phone number. These “creatures of habits” can be more easily identified, say in the case of fraud or terrorism prevention.
Challenges
One potential impediment to leveraging contextual network information involves the privacy and data security issues. Since contextual networks tend to involve personal transactions, companies would like to ensure the highest security to prevent breaches. Various regulatory bodies at different state, country or global levels might have conflicting guidelines on what information is considered private and cannot be used for data analytics.

Another obstacle today is the scalability challenge. Often these contextual networks are many times the size of the current data used for analysis. Contextual networks rely on relationship information, so sampling runs the risk of losing valuable information. James Kobelius of Forrester postulates that the advent of social network analysis will catalyze the need for petabyte-sized data warehouses.4

Another obstacle is the lack of software tools that will generate these graph-based predictors. Types of specialized graph based calculations include community detection, diffusion simulation, vertex similarity, topology and centrality analysis. Open source tools such as Jung and Pajek work well for R&D, while proven commercial tools such as Sonamine are geared toward large scale and production graph mining.

A last obstacle is education. Contextual network predictors are unlike other types of data being captured and analyzed. Enterprises must first capture this new data, and that will require some mind-set shift. Setting up an infrastructure to model the contextual network will require new tools and approaches. Finally, using these contextual networks in predictive analytics that improve the bottom line will require the cooperation of multiple disciplines—analytics, marketing and operations.

Despite obstacles, the time to investigate and leverage these graph-based methods is now. IDAnalytics, a company that provides identity theft protection, has been issued a patent for a system and method for identity-based fraud detection through graph anomaly detection. How will you start adding graph-derived predictors into your predictive analytics?

Dr. Zeller manages the strategic direction of Zementis, a software company focused on predictive analytics and advanced decisioning technology.

In this interview, Dr. Zeller explains how organizations increasingly recognize the value that predictive analytics offers to their business. The complexity of development, integration, and deployment of predictive models, however, is often considered cost-prohibitive for projects. In light of mature open source solutions, open standards, and SOA principles we propose an agile model development life cycle that allows us to quickly leverage predictive analytics in operational environments.

What you will learn:

* Leverage predictive analytics in real-time
*Accelerate time-to-market for decision models
* Reduce cost with SaaS

Transcript:

Bill Cullifer, WOW I am on the phone with Dr. Zeller, CEO of Zementis out of San Diego, California. Dr. Zeller, good afternoon and thanks for agreeing to the call.

Dr. Zeller: Thanks for having me on here. I think excellent educational website. So, I am glad to contribute.

Bill Cullifer, WOW: I appreciate that. You have experience as a computer science fellow and you have been doing predictive analytics for a number of years. You and have a session coming up on the San Diego Computer… Supercomputer Center High Performance Scoring of Healthcare Data as it relates to predictive analytics. Can you give us a couple of walk aways from that session?

Dr. Zeller: Absolutely. It’s going to focus on the ease of deployment for predictive analytics, which really is also the core focus of Zementis and our product line. Taking predictive models that you have developed in various different models and bringing them into a production environment so that any other IT system can easily leverage it, utilize and kind of shorten that time to deployment, time to market predictive analytics.

Bill Cullifer, WOW: And who exactly benefits from that statement? Are we talking about the department itself organizationally, what… where is the return on investment?

Dr. Zeller: Well, the return is really across the enterprise, bringing it into an operational infrastructure and much easier than it has been done in the past. Most of the time they are great tools for data mining, analysis, and model building, but the main hurdle today actually is really the deployment of predictive analytics and…

Bill Cullifer, WOW: Yeah, interesting.

Dr. Zeller: So many companies are not willing or couldn’t spend you know the time and cost associated with such a project. So, it has been I think a limiting factor for predictive analytics in general, just not being able to easily move things into a production environment at a reasonable you know very cost effective way.

Bill Cullifer, WOW: Yeah and how has that changed?

Dr. Zeller: Well, what we are doing is really leveraging the predictive model markup language (PMML), which is the widely known standard in the industry and supported by really the major industry players. We are really focusing on open standards and utilizing that in order to transition from any tool, commercial or open source, into our deployment In addition to that, we actually leverage a new trend that is a very hot topic currently in the industry. It is called computing. So, we leverage for example the Amazon Elastic Compute Cloud to deliver our software as a service at a much, much lower cost than you have ever seen that before.

Bill Cullifer, WOW: Excellent and that’s a subscription model.

Dr. Zeller: Correct yeah. Instead of spending $50000 or $100000 or more on software and hardware licenses, here you can get started at less than a dollar an hour.

Bill Cullifer, WOW: Excellent and where might people go to get more information Dr. Zeller?

Dr. Zeller: Our website, www.zementis.com should really give kind of all the details, you can sign up, you can really get started within five minutes, launch your instance and deploy your models and be in production literally within you know a few hours.

Bill Cullifer, WOW: Excellent. Sounds really good and Dr. Zeller, we certainly appreciate the information and your time today.

Dr. Zeller: Very much my pleasure.

Today s podcast is a continuation of the coverage of the topic Predictive Analytics, To help us better understand the topic from a solution provider point of view, I interviewed Mary Grace Crissey, Global Marketing Manager for Analytics from SAS.  To listen to the entire interview follow the “click to play” link at the bottom of this post.

Transcript of Interview with Mary Grace Crissey

BILL CULLIFER: Greetings WOW Members and Web Professionals everywhere! Bill Cullifer here with the World Organization of Webmasters (WOW) and the WOW Technology Minute. Today’s podcast is a continuation of a series of podcasts on the subject of predictive analytics. To assist us in better understanding the topic from a solution provider point of view, I’m on the phone with Mary Grace Crissey, Global Marketing Manager for Analytics for SAS. Good afternoon Mary Grace and thanks for agreeing to this interview.

MARY GRACE CRISSEY: Well thank you for inviting me.

BILL: You bet. Mary Grace, SAS recently announced the company supercharges predictive analytics. What exactly do you mean when you write that SAS supercharges predictive analytics? And can you explain to the listeners and the viewers of this podcast what predictive analytics means to you?

MARY: Certainly. Well, the word supercharging was tossed in there because it shows my enthusiasm on this topic here.

BILL: All right.

MARY GRACE: I really am a firm believer that you can get some magical power, some turbo-charge and super-boost out of your day to day business, be it making decisions as an executive leader at the company or as a Web professional designing and promoting your webpage. All of these tasks that are commonly done in businesses today can be supercharged if you take advantage of predictive analytics. So when I say predictive analytics I’m talking about looking forward into the future with a strategic mindset. Analytics is a term that represents numbers and metrics and when you want to find patterns, you’re digging through data that you have collected, say of people who visited your website, or demographics of who these customers are, you can look at historical data and there’s a lot of valuable analytic information that you can find. But when you want to look forward and project onto the future, that’s when you’re talking about predictive analytics.

What was once reserved to say an Ivory Tower of the scientific investigations where the academic analyses were punching and forming and computing these number-crunching, data-mining technique, now it’s catching the eyes of successful business leaders. So armed with this latest technology we can all benefit today from crunching word and numbers at the same time. We can automate manual tasks that used to take a lot of people power and actually make informed, fact-based decisions.

BILL: Excellent. Excellent explanation. I’m curious to know Mary Grace, from your point of view, how much of this is preparation and how much of the process, if you will, is technology?

MARY GRACE: Good question because actually the technology and data-mining is sometimes looked at as one little piece of building a predictive model, but SAS has an enterprise intelligence platform that actually spans from the beginning where you collect the data and where you present the information on the webpage, or Web analytics, and where you make the models and actually where you make the decisions. SAS is talking about a solution which truly does entail consultants and experts, people power, behind the software.

BILL: Excellent. Well said. Thank you so much Mary Grace for your time today and for your terrific explanation of predictive analytics. We appreciate it. Bill Cullifer here with the World Organization of Webmasters (WOW), on the phone with Mary Grace Crissey, Global Marketing Manager for Analytics for SAS. Thanks for your time Mary Grace.

MARY GRACE: You’re welcome.