Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
‘How Target Figured Out a Teen Girl Was Pregnant Before Her Father Did’ To begin with, the Internet of Things has facilitated the collection of vast chunks of data, some of which lead to accurate insights when analyzed. In this course, we learnt about the four V’s of big data that are applicable in the world of computing. They include volume, veracity, velocity and variety. Firstly, huge volumes of data can be collected using IoT devices. Additionally, this data comes in multiple forms known as varieties. This is based on the origin of the data. Thirdly, this data can be streamed from one source to the other by connecting multiple devices. This is the concept of velocity. Fourthly, collection of huge loads of data makes analysis challenging which is referred to a veracity of data. This article by Forbes explains how Target was able to predict the pregnancy of a teenage high school girl even before her father had known. Undoubtedly, all stores collect vast loads of data from the customers. Such data include the customers’ names, email addresses, phone numbers, credit card numbers, to name just a few. Additionally, when a customer shops, their stock of goods is saved, and personalized ads are sent to them in case a new product is developed or when the product is on offer. In this particular case, Target made powerful insights about pregnancy. One of the store’s staff noted that pregnant women bought large quantities of unscented lotion instead of the rest, who preferred lotions with a strong scent. Another staff noted that, in the 20th week of pregnancy, women bought large quantities of supplements such as calcium, magnesium, and zinc. Still, during pregnancy, women bought large quantities of scent-free soaps instead of other customers who prefer a strong scent. Moreover, the store’s staff noted that women approaching their delivery dates buy hand sanitizers and washcloths. Therefore, when the store collected data on this teenage girl, it predicted that she was pregnant, and her delivery date was approaching. In summary, the store has 25 items on the checklist that enables it to predict pregnancy. In this scenario, when the store sent a coupon to the teenage girl, the father saw it and became very upset. He walked into one of the Target Stores in Minneapolis and demanded to talk to the manager. He complained to the manager that his daughter, who was only a teenager, had been sent a mail that somehow encouraged pregnancy. The manager apologized on behalf of the store’s staff and even made a call after some days to offer another apology. However, when he made the call, he realized that the father had a somber mood and a relaxed tone, unlike when he visited the store. After a brief conversation, the father said that he had since talked with his daughter, who confirmed that she was pregnant. This case depicts how the Internet of Things has simplified data collection and how this data could be used to make powerful insights. ‘Microwave oven to blame for mystery signal that left astronomers stumped.’ Undoubtedly, data collection can lead to powerful insights when this data is analyzed and processed. However, the process has its considerable flaws that, if not prevented, can lead to inaccurate insights. Bias can be found in both large and small loads of data; however, the risk increases with the increase in the quantity of data. For instance, in this scenario, Australian astronomers had been baffled for more than 17 years on the source of interference that occasionally stroke their telescope. When the interferences began in 1998, the astronomers thought they were caused by lightning. After a thorough examination, they resolved that the interferences came from only 5km away from the telescope. Additionally, they noted that the interferences only happened during the day and only a couple of times in a year. They tested the strength of the signals, which was at about 2.4 GHz. This signal strength was typical of that produced by a microwave. Therefore, since the organization had a microwave used occasionally and only when employees had to warm their lunch and coffee, the scientists resolved to test whether the microwave caused the signals. No interferences were detected when they tested the microwave in a closed door. However, when the door was open, the interferences were detected. Surprisingly, the telescope would only be affected when the dish was pointed in the direction of the microwave. This information helped to demystify a myth that had been previously held for more than 17 years. This case scenario perfectly reflects how data, even though collected accurately, could lead to inaccurate insights. Therefore, people should be cognizant of potential biases that would otherwise compromise the accuracy of the data and the accuracy of the insights. Admittedly, this is not easy going; however, we must try as much as possible to minimize or eliminate bias altogether. In the course, we studied the four main problems encountered in gathering big data. They include; uncertainty of the data, collection of multiple forms of data, it is challenging to analyze among others. For instance, in this case scenario, the astronomers gathered huge loads of data that tried to explain the origin of the interferences. However, they were unable to make accurate analysis since the wrong data had been collected. In a nutshell, for more than 17 years, the misery remained unsolved.