A short introduction to Mahout during SCISR meetup http://bit.ly/scisr

1. Ofer Vugman
May 2012

2. Agenda and such…
 What is ML (Machine Learning)
 ML Common Use Cases
 Mahout Overview
 Algorithms in Mahout
 Mahout Commercial Use
 Mahout Summary

3. What is ML
“Machine Learning is programming
computers to optimize a performance
criterion using example data or past
experience”
 Intro. To Machine Learning by E. Alpaydin

4. ML Common Use Cases
 Recommendation

5. ML Common Use Cases
 Classification

6. ML Common Use Cases
 Clustering

7. ML Common Libraries

8. Mahout Overview – What ?
A mahout is a person who keeps and drives
an elephant

9. Mahout Overview – What ?
 A scalable machine learning library

10. Mahout Overview – What ?
 Began life at 2008 as a subproject of
Apache’s Lucene project
 On 2010 Mahout became a top-level
Apache project in its own right
 Implemented in Java
 Built upon Apache’s Hadoop (Look ! An
Elephant !)

11. Mahout Overview – Why ?
 Many open source ML libraries either:
 Lack community
 Lack documentation and examples
 Lack scalability
 Lack the Apache license
 Are research oriented
 Not well tested
 Not built over existing production quality
libraries

12. Mahout Overview – Why ?
 Scalability
 Scalable to reasonably large datasets (core
algorithms implemented in Map/Reduce,
runnable on Hadoop)
 Scalable to support your business case
(Apache License)
 Scalable community

13. Mahout Overview – Why ?
 Built over existing production quality
libraries

14. Mahout Overview – Use Cases
 Mahout currently supports mainly four
use cases:
1. Recommendation
2. Clustering
3. Classification
4. Frequent Itemset Mining

15. Mahout Overview - Technical
 System Requirements
 Linux (or Cygwin on Windows)
 Java 1.6.x or greater
 Maven 2.0.11 or greater to build the source
code
 Hadoop 0.2 or greater*
* Not all algorithms are implemented to work on Hadoop clusters

16. Algorithms in Mahout
 We’ll focus on one example:
 Collaborative Filtering (Recommenders)
 Yet there are many (many !!) more, you
can find them all on
https://cwiki.apache.org/confluence/dis
play/MAHOUT/Algorithms

17. Algorithms Examples –
Recommendation
 Help users find items they might like
based on historical preferences
 Based on example by Sebastian Schelter in “Distributed Itembased
Collaborative Filtering with Apache Mahout”

18. Algorithms Examples –
Recommendation
Alice 5 1 4
Bob ? 2 5
Peter 4 3 2

19. Algorithms Examples –
Recommendation
 Algorithm
 Neighborhood-based approach
 Works by finding similarly rated items in the
user-item-matrix (e.g. cosine, Pearson-
Correlation, Tanimoto Coefficient)
 Estimates a user's preference towards an
item by looking at his/her preferences
towards similar items

20. Algorithms Examples –
Recommendation
 Prediction: Estimate Bob's preference
towards “The Matrix”
1. Look at all items that
 a) are similar to “The Matrix“
 b) have been rated by Bob
=> “Alien“, “Inception“
2. Estimate the unknown preference with a
weighted sum

21. Algorithms Examples –
Recommendation
 MapReduce phase 1
 Map – Make user the key
(Alice, Matrix, 5) Alice (Matrix, 5)
(Alice, Alien, 1) Alice (Alien, 1)
(Alice, Inception, 4) Alice (Inception, 4)
(Bob, Alien, 2) Bob (Alien, 2)
(Bob, Inception, 5) Bob (Inception, 5)
(Peter, Matrix, 4) Peter (Matrix, 4)
(Peter, Alien, 3) Peter (Alien, 3)
(Peter, Inception, 2) Peter (Inception, 2)

22. Algorithms Examples –
Recommendation
 MapReduce phase 1
 Reduce – Create inverted index
Alice (Matrix, 5)
Alice (Alien, 1)
Alice (Inception, 4) Alice (Matrix, 5) (Alien, 1) (Inception, 4)
Bob (Alien, 2) Bob (Alien, 2) (Inception, 5)
Bob (Inception, 5) Peter(Matrix, 4) (Alien, 3) (Inception, 2)
Peter (Matrix, 4)
Peter (Alien, 3)
Peter (Inception, 2)

23. Algorithms Examples –
Recommendation
 MapReduce phase 2
 Map – Isolate all co-occurred ratings (all
cases where a user rated both items)
Matrix, Alien (5,1)
Matrix, Alien (4,3)
Alice (Matrix, 5) (Alien, 1) (Inception, 4) Alien, Inception (1,4)
Bob (Alien, 2) (Inception, 5) Alien, Inception (2,5)
Peter(Matrix, 4) (Alien, 3) (Inception, 2) Alien, Inception (3,2)
Matrix, Inception (4,2)
Matrix, Inception (5,4)

24. Algorithms Examples –
Recommendation
 MapReduce phase 2
 Reduce – Compute similarities
Matrix, Alien (5,1)
Matrix, Alien (4,3)
Alien, Inception (1,4) Matrix, Alien (-0.47)
Alien, Inception (2,5) Matrix, Inception (0.47)
Alien, Inception (3,2) Alien, Inception(-0.63)
Matrix, Inception (4,2)
Matrix, Inception (5,4)

25. Algorithms Examples –
Recommendation
Alice 5 1 4
Bob 1.5 2 5
Peter 4 3 2

26. Mahout Commercial Use
 Commercial use

27. Mahout Resources
 Mahout website - http://mahout.apache.org/
 Introducing Apache Mahout –
http://www.ibm.com/developerworks/java/lib
rary/j-mahout/
 “Mahout In Action” by Sean Owen and Robin
Anil

28. Mahout Summary
 ML is all over the web today
 Mahout is about scalable machine
learning
 Mahout has functionality for many of
today’s common machine learning tasks
 MapReduce magic in
action

29. Mahout Summary
Thank you and good night