Introduction to NSA data mining

The agencies of the government that is concerned with security, intelligence, and law enforcement are the National security agency (NSA), the Federal Bureau of Investigation (FBI) and the central Intelligence Agency (CIA). Though each agency is separate there come situations where they may need to correspond and exchange information so that they can achieve a breakthrough on a specific case. The people who work in these areas are highly specialists in their field.

The National security agency is seemingly the least known and the most poorly understood government intelligence agency of all the ones in existence. It is a combination of the central security service and a crypto logical organization. This organization protects sensitive information by decrypting foreign intelligence and generating encryption keys to protect American information.

‘Informative assurance’ is the ultimate mission of the NSA. This refers to breaking the codes of foreign intelligence while maintain the protection of the American computer intelligence decryption. The NSA also deals with American computer intelligence and decryption and has made a significant contribution to computer technology around the world. Some interesting facts of the NSA are that the early computers were developed in their office and they are closely affiliated with the U S army in securing military data on bases and in the battlefield.

What is government data mining?

This is gathering existing information from different large locations to generate new and useful information. Algorithms that use techniques from stats and machine learning are employed to search large databases of information automatically. Another name for data mining is Knowledge-Discovery in Databases (KDD). Data mining is quite similar to artificial intelligence and is applicable to a variety of number varying activities.

Data mining is used more frequently in the corporate world to determine the direction of trends and to predict the future. It is used to create models and decision support systems that give people useful information and takes a front line role to combat terrorism. It is speculated that this system was used to help identify the leader in the attacks of 9/11.

Data miners are statisticians and they use techniques with names such as:

• Near-neighbor models
• K-means clustering
• Holdout means clustering
• Holdout method
• K-fold cross validation
• Leave-one-out method

Irrelevant patterns are subtracted by regression techniques therefore leaving useful information only. 

Important terms relevant to NSA data mining

• Bayesian – this refers to a class of interference techniques that tells the likelihood of events that will take place in the future by analyzing and combining probabilities and prior probabilities based on conditional events.
• Spam filtering – this is a type of data mining that brings relevant messages automatically to the surface from a chaotic sea of phishing attempts and Viagra pitches.
• Decision trees – these are used for filtering large amounts of data. Large data passes through here through an entrance node. It then faces a filter at separates the information into different cascades that are determined by different characteristics. An example is information about consumer behavior. This information is likely to be categorized based on demographic factors and age of the customers. Though internet data mining and mining in general  is not all about how impressive graphs and visual representations can be, doing this does help since it is an easier way for bulks of information to be read and because the mind tends to retain such information longer and more efficiently when it is presented in such a way.

Weaknesses in data mining 

It is quite important that when you are learning about something you know the positive and negative effects of it as well as its disadvantages and advantages.  While data mining by the NSA is quite important there are also flaws in the system like any other man made system.  Here are some of the weaknesses that are found in National security agency data mining:

• The accuracy of the results as purported by the NSA is sometimes ‘off’. There are other methods that you can employ when applying various layers of filters and while this is absolutely true the base level of inaccuracy still continue.
• According to the NSA, some internet communications and phone records are captured and are then analyzed to apply a number of indicators to decide if the subject is foreign.
• If there is a normal civilian that knows about algorithms then they will be able to manipulate the results. These may be private but it does not mean that people are completely unable to access it. As there is an increase in the revelation of the program so is the information associated with it once the results are manipulated.
• It was said that the government used US data mining to prevent attacks from terrorist on the nation. So, if they were able to use various methods to gather information beforehand that would assume that these same methods may be used to obtain private information from citizens.
•  For means of comparisons and analogy, if it was the case where the government was really monitoring homes privately then they would be able to realize just how many cases of domestic violence takes place on a daily basis.  On the other hand, if they are really able to do this based on what they have done to obtain information in the past why is that they are doing nothing about it?