Data Mining Movie Survey

In November 2002, the SQL Server Data Mining team created a survey asking Microsofties questions about their movie viewing behavior, their demographics, and their favorite hobbies, movies, actors, and directors.  Almost 3200 responses were gathered and $1000 in prizes were distributed to over 40 lucky winners with a $500 Grand Prize. The results of the survey were used to exercise the advanced data mining capabilities of SQL Server Yukon. Here you can find out about how we mined the data, and actually play with some of the mining models yourself to learn more about Microsofties and how they like their movies….

• Survey Questions – Read the questions that were asked in the survey.
• Data Gathering – Learn how we gathered and transformed the data collected for the survey
• Mining – See how the data was mined and the models that were created.
• Favorites – Some quick “Top 10” lists for Microsoft.
• The Forests – Interactively explore trees describing different aspects of Microsofties’ lives and movie watching behavior that were created using the Microsoft Decision Trees algorithms
• What Differentiates – Interactively explore models that can tell you the differences between different types of Microsofties – for example, “What’s the difference between homeowners and renters?” or “What’s the difference between men and women?” These models were created using the Microsoft Naïve Bayes algorithm.