330 likes | 466 Vues
Data Mining. - A Powerful Computing Technology. Department of Computer Science Wayne State University. Road Map. Overview Recommender Systems Clustering Classification Association Analysis PageRank Social Networks. Different Forms of Data. Text Data. Different Forms of Data.
E N D
Data Mining - A Powerful Computing Technology Department of Computer Science Wayne State University
Road Map • Overview • Recommender Systems • Clustering • Classification • Association Analysis • PageRank • Social Networks
Different Forms of Data • Text Data
Different Forms of Data • Image Data
Different Forms of Data • Video Data
Different Forms of Data • Network Data
Why Data Mining is Important? • Difficulty of identifying patterns in big data. • Extracting only WANTED data within a short time. We are drowning in data, but starving for knowledge!
How Data Mining can help? • We do not care if GOOGLE has more than billion web pages. • We only care about the information that is useful for us.
What is Data Mining • The analysis of data to extract useful patterns or information from a large data collection. • Also known as: Knowledge Discovery in Databases • Learn More: http://en.wikipedia.org/wiki/Data_mining Automated Analysis of Massive Data
Data Miner • An educational tool that teaches you Data Mining techniques. • Consists of two basic parts such that, • Demonstration • Explains how to work with the interactive part. • Interactive part • Teaching data mining through user interaction.
Recommender Systems • Goal: present information items that are likely to be of interest to the user. • Lots of online products, books, movies, etc. • Reduce my choices…please!!!! • Learn More: http://pespmc1.vub.ac.be/collfilt.html
Recommender Systems • Netflix Recommender System
Do you watch movies using Then you might like So on you might like these too Or may be you like If you have watched this movie This might catch your interest too
Amazon Recommender System • Amazon Recommender System
Data Miner - Recommender System • Recommendation based on content
Finding a Friend With Similar Taste YOU See what they like Measure the similarity Select your Neighbors
Cluster Analysis • Cluster: • A collection of data objects • Cluster Analysis: • Grouping some given objects with similar attributes. • Similar (or related) to one another within the same group • Dissimilar (or unrelated) to the objects in other groups • Learn More: http://home.dei.polimi.it/matteucc/Clustering/tutorial_html
Cluster Analysis • Data Set: • Clusters: Flowers Fruit
Clustering • Now you have seen Flowers and Fruits visually. • So to which cluster, would you add this object? Flowers Fruit Yes, to FRUIT!!
Classification • Assigning given items to a known class which have items with similar attributes. • Explains through Decision Trees.
Classification • PURE Classification. • Each branch contains animals belong to a single CLASS.
Classification • You have learned what is Mammal and what is Bird. • Can you tell what is this? Yes, this is indeed a BIRD!!
Association Analysis • Discover interesting relationships in a set of transactions. • Understand relationships between items. E.g. • If a customers buys shoes, then 10% of the times they also buys socks. • 60% of all shoppers will buy bread when they also purchase a pint of milk.
Association Analysis • Items: • Transactions:
PageRank • Links from popular and related web sites increases the popularity of the given web site. Yahoo Amazon Pillsbury YouTube Billboard Pandora Dominos Pizza Crayola Pizza Hut Danskin Shelfari
Search Results • When searching on Google, it will list web sites related to the input text according to their importance.
Social Networks • Social networking websites allow users to be part of a virtual community. E.g. Facebook, Twitter, MySpace • They provide users with simple tools to create a custom profile with text and pictures. • Users can share their lives with other people through these networks.
Social Networks • Learn More: • http://en.wikipedia.org/wiki/Social_network • http://pc.net/glossary/definition/socialnetworking
Thank You !! Enjoy the Day…