Diaper and beer

The case of the market basket analysis in the United States that "as for the person who buys the diaper there isa tendency which buysthe beer". From the nineties middle applying on 2000 age beginning, being the media, and lecture etc. you talked wellthe word,data mining you made concept suddenly famous.

Generally "the result of analyzing the sale data with the major supermarket chain of the United States, as for the customer it understood that it is the tendency which buys the diaper and the beer together. The result of investigation, with the home where the child is as for the mother in order to buy the paper diaper which be bulkies you asked to the father, the father who comes to the store next had purchased the can beer. Then arranging this two, when it displays, it is known with the contents that the sale rose ".

Directly, it assumes that was published to "the Wall Street journal" 1992 December 23rd the article, "Supercomputer Manage Holiday Stock" is beginning. With this article "as for this store, when we assume, that a certain person at 5 o'clock in the afternoon bought the paper diaper, you announced at the city of the American Nakanishi section that next the fact that the possibility half a dozen of buying the beer it is large was discovered", but while being conveyed, name of the store became seven - became eleven, or day and time was Thursday, was weekend and/or with the various variations appeared. Because of this, whether or not the case which really is doubt apparent it reached the point where it is done, was called "legend" often.

According to the place where the president of consulting company MindMeld and the Thomas A Bliss cock (Dr. Thomas A. Blischok) person of CEO talked in 2002, around 1992 the same person is the vice president of the business consulting charge of NCR, obtaining the transaction data (market basket data of 120 ten thousand or more) from the register of 25 stores of the retail store chain, Osco Drugs it is analysis that it is original to do.

As a result, you say "that the fact that was discovered between 7 o'clock, the consumer buys the diaper and the beer from 5 o'clock in the afternoon,". This is examined also one time and the callous the first case of the data mining in the sense that the discernment which does not grind could be obtained, you can say that it is, but as for the fact that Osco arranged the diaper and the beer in the same counter, on the basis of this knowledge you say that it is not.

Activity of the Bliss cock and othersbeforefirst algorithm (Apriori) of association analysis is utilized, is thought that perhaps, in the story, with actual analysis the SQL technician exists inside the transaction data, defining "affinity", you say that it drew up

In this research simultaneously "the juice and weir setting agent" "cosmetics and greeting guard" "candy and the greeting card" and so on, it is the combination where 30 differs that it is specific. However, as for correlation of the diaper and the beer there is also a theory which we assume that to tell the truth it was not supported, this time.

As for the nineties first half as for the word, data mining it was not known yet, it was immature even technically and either application method to business had not completely understood. Among those the story "of the diaper and the beer" it explained it was superior well concerning the meaning and possibility of the data mining, you probably can say that "legend" was.