A Data Warehouse - Assignment Example

Answer A data warehouse is the “collection of integrated, oriented databases designed to supply information required for decision making” (Inmon, 1992). For the success of any business, it is necessary to keep in its data warehouse, the most update information. However, various types of organizations have varied urgency for a particular type of data and information. Ranked in the order of the need of the most up to date information to the company needing the least time-sensitive information is: 1. Hannah’s Entertainment: Being in a wide range of business from hotels, to casinos, it is imperative for Hannah to use real-time analysis. As soon as customer interact with slot machine, or buy meal, predictive analysis can be used to offer them rewards. 2. Einstein/ Noah Bagel: Einstein / Noah Bagel want to use the data to maximize cross-selling opportunity. This is possible if it has real time data in its warehouse. The daily cross-selling opportunities can be informed to the managers if the data warehouse contains the data for daily transactions. 3. AFC Enterprises: AFC needs to track customer-specific information to anticipate, and manipulate consumer behavior. The data warehouse has to be monitored with every transaction with the customer. 4. Red Robin International: Red Robin International wants to use its data mining tools to analyze the effects of its promotions, and menu lists. These decisions have to be taken with a low frequency, and the data warehouse can be updated with a low frequency. 5. Pizzeria Uno: It uses data mining tools to identify 80% of the most profitable customers. These customers remain the same over a long period of time, so there is little relevance of up to data warehouse in this case. Answer 2. Dimensions are the basis for new methodology on requirement analysis. They contain static details about the transactions in fact table (Powell, 2005). For the hotel segment of Hannah Entertainment, it is necessary to analyze hotel occupancy of the rooms in various centers of the hotel, in different seasons. Hannah will like to analyze sales by hotels, by types of rooms occupied, by customer history, and by various seasons. So, the critical dimensions of a data warehouse for hotels are hotel, type of the room, customer, and season. For a restaurant, the various dimensions which shall be the basis of analysis include: restaurant location in a city, region, menu items, time periods such as day, afternoon, or night, customers, and items ordered together. For a gaming casino, the dimensions will be the player, the machines, the various games, time, and location. Answer 3. Gathering sensitive information such as the name, and address for a data warehouse is one of the most important tasks for creating a data warehouse. There are various sources from where, such information can be obtained. One of the probable sources, where AFC could gather this data is through a customer feedback system, after they have completed their meal. This feedback system might ask the customer for filling forms, and ask for customer name, address, and other details. Another likely source of this information can be purchasing this information from some third-party vendor. The information from a third party vendor might be unethical if the vendor has not taken the permission from the customer to use his information for commercial purpose. Whatever be the method of collection of information, it is necessary, that AFC does not sell it to somebody else. Answer 4. The breakfast cereal aisle in the grocery store was categorized by availability of a number of brands, and the number of choices for each brand. The mild taste and lower priced options were on the lower shelves, and were harder to reach. The well known and higher priced brands were put on shelves visible at a straight line from the normal eye. These also covered the maximum space on the shelves. Children’s premium breakfast cereals are kept at lower shelves at their eye levels. I also found that the candy and breakfast cereals were kept on the opposite sides of the same aisles. The placement of cereal breakfast on shelves is a major decision taken by grocery stores. The decisions to place a particular cereal at a particular position depends on factors like the price, the demand for the brand, the relation between various categories of items, the types of customer buying a particular product, the brand awareness of the cereal, and inventory cost for a particular type of cereal. A very appropriate source of this information is a data warehouse containing the data from their POS transaction system. Data mining tools may be used to understand the direct elasticity, and cross elasticity of a particular product. Techniques like association rule mining can be used to find out relations between two products, categories, subcategories. This association rule can then be used to design the best layout plan of shelf positioning for maximum profits. Dynamic Mining association can be used to discover changes in customer buying behavior over time (Shenoy et. Al, 2002). Answer 5. In order to predict the toppings by the time of the day and the week, it is needed to have an operational data warehouse. These types of data warehouses contain the summarized, historical view of data (Humphries et. Al., 1999). The data warehouse shall contain the history of past sale of a topping classified according to the time, the season and the customer. Data mining techniques can then be used to identify relations between various time periods, season, customer preferences and toppings. The data warehouse will be a useful one, if one is able to analyze the sales of particular topping with respect to various time periods, like the afternoon, night, and morning. It is also needed to analyze the sales of a particular topping with respect to various seasons. Besides, we should also be able to analyze customer history, and identify customer preference. The dimensions of the data warehouse shall be: Toppings, Time periods, Seasons, and Customer. In order to increase the sale of a particular topping, it is necessary to understand, the time period, and season when the sales are maximum for the topping. Special promotion schemes can be then implemented for the topping during that period. If I am able to see the group of toppings ordered together, and want to increase the sales of one of them, then promotional schemes can be made for other toppings, as well. References Humphries, M., Hawkins M. W., Dy M.C. (199) Data Warehousing: Architecture and Implementation. Prentice Hall. Inmon, William, H. (1992) Buidling the data warehouse. QED technical Publishing Group. Powell G (2005). Beginning Database Design. Wiley Publishing Company. Shenoy, Deepa, P., Srividya M, Kavitha, Venugopal K.R. & Patnaik, L.M. (2002) Incremental Mining of association rules using genetic algorithm”. Proceedings from Intl. Conf on high Performance Computing (Asia). Read More