Welcome to the Wiki of the course '''Information Retrieval''' in the winter term 2012/2013
The course is given by Prof. Dr. Hannah Bast and assisted by Florian Bäurle. It takes place every Wednesday from 4:15pm until 5:45pm, in seminar room 00-010/14 in building 101. The first lecture is on Wednesday, October 24, 2012 and the last lecture is on Wednesday, February 13, 2013. There will be no lecture on Wednesday, December 26, 2012 and Wednesday, January 2, 2013 (christmas break) and on TODO. That is 14 lectures altogether.
The tutors for this course are Mirko Brodesser (mirko.brodesser@gmail.com), Eugen Sawin (esawin@gmail.com), Jonas Sternisko (js174@mars.uni-freiburg.de), Ragavan Natarajan (n.ragav@gmail.com). The administrator of the supporting systems (Daphne, SVN, Forum, Jenkins) is Axel Lehmann.
Important Links
Here is the link to Daphne (our course management system).
Here is the link to the Forum (for annoucements and questions).
Here is the Wiki page of a similar lecture I have given three years ago: Search Engines, WS 2009/2010.
Here is a Java Code Style Profile for Eclipse: Eclipse Java Code Style Profile.
Here is a C++ Code Style Profile for Eclipse: Eclipse C++ Code Style Profile.
Lecture Slides, Video Recordings, Exercise Sheets, and Code
Remark about the video recordings: we have prepared videos in different formats, in particular: .m4v, .mp4, and .wmv. If you click on one of the links "Video recording" you get that format of which our script thinks that it best fits your device / operating system / browser. If you explicitly want a particular format, simply copy the link address and add the desired extension by hand.
Lecture 1, Wednesday, October 24, 2012 (Introduction, Organizational, Inverted Index): Video recording, Slides, Exercise Sheet 1, Code from the lecture + design suggestion for the exercise sheet, Example collection (1.1M sentences from Wikipedia about scientists), Master Solutions 1.
Lecture 2, Wednesday, October 31, 2012 (Vector Space Model, Ranking, Precision/Recall): Video recording, Slides, Exercise Sheet 2, Code design suggestion for the exercise sheet, Master Solutions 2.
Lecture 3, Wednesday, November 7, 2012 (List intersection: fancy algorithms + lower bounds): Video recording, Slides, Exercise Sheet 3, Code from the lecture, Table for your list intersection results, Master Solutions 3.
Lecture 4, Wednesday, November 14, 2012 (Compression and entropy): Video recording, Slides, Exercise Sheet 4, Master Solutions 4.
Lecture 5, Wednesday, November 21, 2012 (Prefix search, error-tolerant search): Video recording, Slides, Exercise Sheet 5, Code design suggestion for the exercise sheet, Vocabulary (291.846 words) and Queries (1.000 words) for your experiments, Table for your approximate word matching results, Master Solutions 5.
Lecture 6, Wednesday, November 28, 2012 (Error-tolerant prefix search, web application): Video recording, Slides, Exercise Sheet 6, Code from the lecture + design suggestion for the exercise sheet, Master Solutions 6.
Lecture 7, Wednesday, December 5, 2012 (PHP, Cross-Scripting, Cookies, UTF-8): Video recording, Slides, Exercise Sheet 7, Files from the lecture, Example collection with UTF-8 errors, Master Solutions 7.
Lecture 8, Wednesday, December 12, 2012 (Synomyms, Latent Semantic Indexing): Video recording, Slides, Exercise Sheet 8, Master Solutions 8.
Lecture 9, Wednesday, December 12, 2012 (Clustering, k-means): Video recording, Slides, Exercise Sheet 9, Code design suggestion for the exercise sheet, Table for your clustering results, Master Solutions 9, Alwis (Windows installer for the PLSI demo, virus-free).
Lecture 10, Wednesday, January 16, 2013 (Naive Bayes Classification): Video recording, Slides, Exercise Sheet 10, Code design suggestion for the exercise sheet, Our Wikipedia documents, one line per document now + with labels, Table for your Naive Bayes results.
Lecture 11, Wednesday, January 23, 2013 (Support Vector Machines): Video recording, Slides, Exercise Sheet 11, Our unique Wikipedia documents with labels, only two classes, Table for your SVM vs. Naive Bayes results, SVM Light software (all still preliminary).