{[ promptMessage ]}

Bookmark it

{[ promptMessage ]}

assoc-rules1-4

# assoc-rules1-4 - Association Rules Market Baskets Frequent...

This preview shows pages 1–11. Sign up to view the full content.

1 Association Rules Market Baskets Frequent Itemsets A-Priori Algorithm

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
2 The Market-Basket Model A large set of  items , e.g., things sold in  a supermarket. A large set of  baskets , each of which is  a small set of the items, e.g., the things  one customer buys on one day.
3 Market-Baskets – (2) Really a general many-many mapping  (association) between two kinds of  things. But we ask about connections among  “items,” not “baskets.” The technology focuses on  common  events , not rare events (“long tail”).

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
4 Support Simplest question: find sets of items  that appear “frequently” in the baskets. Support   for itemset  I   = the number of  baskets containing all items in  I . Sometimes given as a percentage.  Given a  support   threshold   s , sets of  items that appear in at least  s   baskets  are called  frequent itemsets .
5 Example : Frequent Itemsets Items={milk, coke, pepsi, beer, juice}. Support = 3 baskets. B 1  = {m, c, b} B 2  = {m, p, j} B 3  = {m, b} B 4  = {c, j} B 5  = {m, p, b} B 6  = {m, c, b, j} B 7  = {c, b, j} B 8  = {b, c} Frequent itemsets: {m}, {c}, {b}, {j}, , {b,c} , {c,j}. {m,b}

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
6 Applications – (1) Items  = products;  baskets  = sets of  products someone bought in one trip to  the store. Example   application : given that many  people buy beer and diapers together: Run a sale on diapers; raise price of beer. Only useful if many buy diapers & beer.
7 Applications – (2) Baskets  = sentences;  items  =  documents containing those sentences. Items that appear together too often  could represent plagiarism. Notice items do not have to be “in”  baskets.

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
8 Applications – (3) Baskets  = Web pages;  items  = words. Unusual words appearing together in a  large number of documents, e.g., “Brad”  and “Angelina,” may indicate an  interesting relationship.
9 Aside : Words on the Web Many Web-mining applications involve  words. 1. Cluster pages by their topic, e.g., sports. 2. Find useful blogs, versus nonsense. 3. Determine the sentiment (positive or  negative) of comments. 4. Partition pages retrieved from an  ambiguous query, e.g., “jaguar.”

This preview has intentionally blurred sections. Sign up to view the full version.

View Full Document
10 Words – (2) Here’s everything I know about  computational linguistics.
This is the end of the preview. Sign up to access the rest of the document.

{[ snackBarMessage ]}

### Page1 / 37

assoc-rules1-4 - Association Rules Market Baskets Frequent...

This preview shows document pages 1 - 11. Sign up to view the full document.

View Full Document
Ask a homework question - tutors are online