by Chris S
The article I chose touches on how tax authorities are challenged with collecting and identifying businesses that do not pay their correct taxes. Believe it or not, businesses and everyday people cheat on their taxes, whether it be for a higher refund or to elude from paying more money back to the government. Of all agencies, the IRS is stated to have limited amounts of resources when it comes to detecting fraud. To try and combat this, a screening system using data mining techniques was created in order to try and “value-added tax” reports, or any other type of suspicious activity. Once these particular activities are recognized, then they can be further audited and have proven effective.
I believe this relates to our class in a number of ways. Obviously the government is using a vast database to try and recognize the fraudulent activities and stop them. Data mining is basically analyzing data from a variety of different approaches and summarizing the findings into a useful set of information. If anyone were to get into the database field, chances are there will be situations where people will have to do these kinds of tasks.
The whole data mining subject reminded me of another story I came across about how a grocery store used data mining to their advantage. It may not be 100% accurate, but from what I remember was that it turned out that on weekends, the sale of diapers and beer were above average. Apparently on the weekends, the father would more often be sent to the store to buy some diapers, and occasionally buy some beer as well. After noticing this type of behavior, the grocery stores relocated the diaper section closer to the alcohol section, thus creating more sales.
Wu, Roung-Shiunn, C.S Ou, Hui-ying Lin, She-I Chang, and David Yen. “Using Data Mining Technique to Enhance Tax Evasion Detection Performance.” Expert Systems with Applications, 39.10 (2012): 8769-8777.