Striking Gold with Data Mining, Analysis Software
November 21, 2006
Technology advances relieve the burden of viewing hours of video images. A similar situation exists for law enforcement and security professionals when reviewing and analyzing recorded security video for specific incidents. Granted, digital recorders have made the procedure faster and easier than having to search through tapes from time lapse VCRs but I believe the recent development and implementation of data mining and processing software is one of the most notable advancements to be introduced to the field of video security technology.
As the name suggests, data mining is the process of sifting through volumes of data looking for the nuggets you have specified in your search criteria. For instance, after the theft of an employee’s purse, building security management can query the system to pull up all incidents of entrance/exit doors being accessed within a certain time frame, along with the accompanying vide
New Name for an Old ConceptThis example illustrates the most basic form of data mining and reminds us the concept is not new, just easier because of digital technology advancements in the security industry and the subsequent ability to store, process and retrieve the various databases (i.e. video surveillance, access control, biometrics, POS, ATM, HR, etc.) in a data warehouse. As discussed in a previous article, data storage or data warehousing on NVRs (network video recorders) or larger NAS (network attached servers) devices allow for the process of centralized data management and retrieval, which is needed to maximize user access and retrieval. Data analysis software then enables user to conveniently access the data.
Most NVRs available today come complete with basic data analysis software with the feature advantage of being able to program and access the functionality from anywhere on the network. As well, the ability to “drill down” offers access to more detailed data and summary information. Using the above example, security management can drill down the data to separate employees from visitors. This is known as “pulling” the information. On a more sophisticated level, data analysis software can be the foundation for a pro-active security-centric business intelligence system. By programming select triggers and queries, users can pre-select potentially problematic situations and record them for later analysis or review in the event of an incident. This is known as “pushing” the information.
Pro-active security and business intelligence for gaming and retail data mining, whether pushing or pulling, is applicable to all areas in the field of security and surveillance. It is however perhaps most beneficial and most predominantly employed in the casino and retail marketplace because of the networked integration of several digital systems and databases within the overall system.
All data of a casino operation, from the gaming floor to the parking lot to the hotel to the restaurants to human resources is warehoused in a multi-dimensional database system. Using the casino’s digital surveillance system, casino operators can set pre-determined triggers of potential loss events in their table game, slot, cage and back of house operations. When an event occurs that matches pre-defined triggers, such as the issuance or redemption of markers at the casino cage or tables, slot jackpot payouts or slot machine hopper fills, the system software will archive the transaction data along with multiple video feeds for each event. In addition to saving time and energy when reviewing material, the data analysis software can also help to identify trends such as scams, employee “sweetheart” deals or suspect behavior patterns.
Data mining of POS transactions has been used extensively in the retail industry for creating marketing promotions, identifying buying trends and developing new product. Now however, with advances in security and surveillance technology and the use of a common digital platform, integration of digital video surveillance with POS (point of sale) or point of use terminal data offers an added application by providing a pro-active loss prevention tool that is quickly starting to be viewed as a must have. Using data mining and analysis software, exception reports that include surveillance video and accompanying POS data can be generated based on the company’s business logic. For instance, if the POS cash drawer is open longer than two minutes, the system will create an exception report with the time, date, employee, cash register, and revenue center along with the video of the event.