Predicting Missing Data

by Kevin S
It may be difficult to translate data in databases which are missing values into useful information. In the article “Prior Knowledge: A Predictive Database For Developers”, Alex Williams discusses how the Prior Knowledge company has decided to tackle blank data. They have released a new Veritable API which, in short, looks at the stored data and intelligently predicts and fills in any blank areas. The hope is that this new technology will allow database developers deploy new applications which can predict the values of these blanks and help turn their data into useful information.

One example included in the article states that (using the new API) “Prior Knowledge worked with a retailer to determine customer purchasing patterns.” (Williams, 2012)

This new API focuses on entity attributes which are left blank in databases. While we haven’t discussed any solutions or data validation for such an occurrence yet, I think it is good to be aware of options available. The ability to automatically analyze the relationship between entity’s with this API to find patterns is impressive as well.

This article serves as an insight to new technology in the field. And as a College of Business student, I have learned the great importance of data integrity when it comes to statistical analysis. This new technology,
which looks at the heart of data to compute patterns, is a prime example of how the world is using computers today in business to increase efficiency (by automatically patching “missing data”) and productivity (by freeing an employee from tedious work) in the workplace.

Williams, A. (2012, September 11). Prior knowledge: A predictive database for developers. Retrieved from http://techcrunch.com/2012/09/11/prior-knowledge-a-predictive-database-for-developers/

3 thoughts on “Predicting Missing Data”

  1. I think it will takes sometime to develop a software that can predict missing data. However, after looking through a few websites, there is a suggestion on how we can predict missing data by using mining algorithm. An example would be Microsoft Decision Trees Algorithm. Basically, it will look at the past data and make prediction and when there is new data, it would compare to see if the predicted data correlated. You can find more information about it here: http://technet.microsoft.com/en-us/library/ms175312.aspx

    http://technet.microsoft.com/en-us/library/ms175312.aspx

  2. I agree with the fact that this article didn’t really pertain to this weeks topic in class but it was a great article regardless. The fact that the API find the patterns for the users is incredible because it doesn’t require the user to have prior understanding of linear congressional analysis, but just the basics of the traditional databases they are already accustomed to.

  3. Data with missing values can definitely create problems for an organization in the long run. It can be frustrating to have a piece of data that doesn’t fit into the diagrams. Thankfully, technology such as the intelligent API makes it much easier for database administrators to manage the server and organize the messy filter.

Comments are closed.