A Look Into the Mind: WiseWindow Leverages Open Source, Cloud Computing to Gauge Opinions

WiseWindow Inc.’s solutions are used to analyze millions of online consumer opinions, helping marketers assess their competitive position and understand what subjects people care about. INTERNET TELEPHONY recently spoke with Rajiv Dulepet, the founder and CTO, and Marshall Toplansky, the president, of WiseWindow, about the company’s Mass Opinion Business Intelligence service, their involvement with Caltech and how open source relates to all this.

What is Mass Opinion Business Intelligence.

T & D: Mass Opinion Business Intelligence is WiseWindow’s monitoring service that gives marketers their products’ real-time share of opinion, analyzes which subjects are dominating the opinions within their product categories, shows who is providing the largest share of opinions and uncovers unmet needs in the category.

Who uses Mass Opinion Business Intelligence today and how?

T & D: Any company that is looking to track online consumer opinion and develop competitive marketing strategies is a logical user of MOBI.

Our clients are confidential; however, here are four examples of how it is being used and by whom:

• A large athletic shoe manufacturer is using WiseWindow to understand who and why people are wearing different brands of fitness shoes.
• A major automobile company is using WiseWindow to understand whether people are actually changing their attitudes and behavior regarding their brand as a result of its marketing campaign.
• A leading healthcare company is using WiseWindow to understand which aspects of healthcare reform are building or losing momentum among the public.
• A major media company is using WiseWindow to rate the popularity of the more than 12 million bands in the country, region by region.

What other interesting potential uses are there for Mass Opinion Business Intelligence?

T & D: For every category we have done work in so far, we have found a fascinating correlation between share of opinions (in some form, whether positive or negative) and the actual share of market (revenue or units) that a brand or product has in the marketplace. As a result of this correlation, we are able to develop predictive demand forecasts for customers. The ability to factor in real-time market opinions into the demand forecasting process is a significant breakthrough in enterprise planning. It has profound implications for manufacturing, purchasing and asset management in an enterprise, not only for marketing and sales.

What is your pricing strategy relative to Mass Opinion Business Intelligence?

T & D: WiseWindow and its value-added partners make money by charging for subscriptions to its data as well as via custom analysis of the data we track.

What does Mass Opinion Business Intelligence have to do with cloud computing?

T & D: Cloud computing is a significant enabling technology for MOBI. Finding opinions that are relevant to a category, and conducting natural language and sentiment analysis, is very computing-resource intense, as is the storage and retrieval of the data that is generated. MOBI distributes the crawling, analysis and database functions across the cloud, which gives us the flexibility to scale usage up and down as required.

I understand Rajiv has been named advisor and architect for a new open source project funded by the National Institute of Health and executed by Caltech. What is the goal of this effort and what exactly does it have to do with open source?

Dulepet: The goal of this effort is to provide scientists an open platform for bio-medical research where they can share analytical applications and data with their colleagues. In addition, we want to provide private companies (like pharmaceutical research firms) with a platform they can use for large-scale analysis and yet maintain their data in a confidential manner. The ultimate goal is to speed up bio-medical research and, where appropriate, enhance collaboration within the research community.

Earlier this decade you served as a visiting scholar at the Stanford School of Management and Engineering, spearheading the development of U.S. presidential prediction analysis. How are you leveraging that experience for your work on Mass Opinion Business Intelligence and on the Caltech effort?

Dulepet: I was involved with Stanford from 2004 to 2006, in addition to doing work on the 2008 election. This experience was the first time anyone was able to leverage unprompted consumer opinion to determine the outcome of elections. At the time, people relied solely on polls to determine the outcome of elections. Polls are biased by the act of asking people their opinions as well as by the selection of the samples used. Mass opinion, by contrast, passively observes what people are saying. The large number of people who express opinions eliminates the need to do ‘select representative’ samples. We found that these opinions were highly predictive of results.

The use of natural language processing and sentiment analysis to analyze mass opinions for the elections was the core analytical process that we have harnessed for MOBI and WiseWindow. The use of the cloud was required to be able to process the enormous number of expressed opinions.

How is your work at Caltech complementary to your development of Mass Opinion Business Intelligence for WiseWindow?

Dulepet: In addition to the cloud computing resource that is part of our Caltech framework, we will be developing an online community of bio-medical researchers, who will express their opinions on various research subjects. We will use MOBI to analyze and track those opinions. IT

» Internet Telephony Magazine Table of Contents