· Ila Gokarn was chosen as a SAS Student Ambassador in December 2014, one of 15 worldwide and the only one from Singapore.
· She will be presenting her research on the different algorithms that can extract significant words to help in prediction and clustering of data records at the SAS Global Forum 2015 in April.
SMU School of Information Systems final-year student and SAS Scholar Ila Gokarn has been selected as a SAS Student Ambassador, one of 15 from around the world, and the only one from Singapore.
The SAS Student Ambassador Program is a competitive program that recognises and supports students who use SAS technologies in innovative ways that benefit their respective industries and fields of study.
This is the second time that a SMU student has been selected as a SAS Student Ambassador. In 2013, Cally Ong, Master of IT in Business (Analytics) student and a SAS Scholar, was selected.
Ila will be presenting her paper ‘Understanding Characteristics of Insider Threats using Feature Extraction’ at the SAS Global Forum 2015 to be held in Dallas, Texas, USA on 26-29 April 2015. Over 3,000 SAS users from more than 35 countries are expected to attend the Forum.
Said Ila, “The conference is a fantastic way to learn more about research being done currently in the field of analytics. With professionals, academics and analytics experts attending and sharing their work, it is also an opportunity to grow both my scholarly and professional networks.”
Essentially, Ila’s research looked into the different algorithms that can extract significant words to help in prediction and clustering of data records. The findings of this research can be applied to any domain whose datasets have a text field.
On the motivation for the research, IIa explained, “During my internship at SAS Institute, I was experimenting on a very large dataset related to fraud – in specific, insider threats – when I found that the data was mostly text, and not adequately structured for in-depth analysis. I wanted to do some data prediction and clustering to find out characteristics common to insider threats in order to be able to predict such threats in the future.
“My mentor at SAS suggested I find a way to extract meaningful words from the text I had, which could help in prediction and clustering. While this can be done using SAS Text Miner software, we were curious if there were other algorithms which could theoretically give us better performance and accuracy. This was the initial motivation for the work, which quickly evolved into an independent research project at SAS.”
Ila then devised a set of experiments to evaluate theoretical algorithms to see which would give the best results during data prediction. She worked on five different algorithms which can all be used for different types of textual data. The experiments were all written in Base SAS and executed using SAS software. The results of the experiments were encouraging and conclusive, leading her to write the paper which she submitted to the SAS Global Forum.
[Featured photo: As one of 15 SAS Student Ambassadors selected worldwide, Ila will be presenting her research on Understanding Characteristics of Insider Threats using Feature Extraction at the SAS Global Forum 2015 in April.]