About: This auction data is gathered from real auctions on ebay.com, crawled by Yuan Wang at University of Wisconsin. This is a somewhat-random, somewhat-edited sample of the data. The Buy_Price and Bids elements were synthetically generated (for a CS145 class project), but all other data is real. The data is organized into files items-*.xml, where each items-*.xml file contains information about 500 auctions. Each auction corresponds to one Item element, which gives complete information about the auction. The files conform to the DTD given in the file data-xml/items.dtd, reproduced here: The meaning of each element and attribute is explained below: ItemID (attribute): An identifier unique across all items. Name: A short item description used as the auction's title. Category: A category to which the item belongs. An item may belong to multiple categories. Currently: The current highest bid. This amount is always equal to the amount of the highest bid, or First_Bid if there are no bids. Buy_Price: This element was synthetically generated. First_Bid: The minimum qualifying first-bid amount, as determined by the seller before the auction starts. It does not mean there is a bid at all. Quantity: The number of copies of the item up for sale. Usually this number is 1, although some auctions are for multiple copies. In such auctions, each bidder may bid on more than 1 copy, and there may be multiple winners. (Note that auction winners are not encoded in our data. It is up to you to determine winners.) Assumed 1 if missing. Number_of_Bids: Number of Bids/Bid elements, each corresponding to a bid. Bids: This element was synthetically generated. Bids/Bid/Bidder: Attribute UserID uniquely identifies a user. Attribute Rating is the user's rating. Note that a user may be a bidder in one auction and a seller in another. However, his Rating, Location, and Country information are the same wherever he appears in our data (which reflects a snapshot in time). Note this implies that UserID's with missing location or country information cannot be sellers in another auction. Bids/Bid/Time: The time the bid was placed. Note that bids must be placed after the auction starts and before it ends. A user may bid on an item multiple times, but not at the same time. Bids/Bid/Amount: Bid amount. If bid quantity is greater than 1, this is the price per copy. Bids/Bid/Quantity: The number of copies bid on. Must be less than or equal to the number of copies up for auction. Assumed 1 if missing. Location: The seller's location information (e.g., city, state). See comment under Bids/Bid/Bidder. Country: Seller's country. See comment under Bids/Bid/Bidder. Started: Auction start time. Ends: Auction end time. If this is in the past with respect to the current system time, the auction is closed. If in the future, the auction is still open. Seller: Attributes give the seller's UserID and rating. Description: The item's full description. All money values are in the form $x,xxx.xx and are in US dollars. All times are in 24-hour format. See the actual data for the exact time format. The auctions in the data set range in time from November to December of 2001. Both open and closed auctions are included, and it is up to you to determine which auctions are still open based on the current system time, taken to be Dec. 20, 2001 00:00:01. Times in the data are consistent with the current system time, so all bid times and auction start times are earlier.