Web Mining

Teachers

• Mauro Brunato - E-mail brunato@disi.unitn.it
  To fix an appointment, please check my availability at http://disi.unitn.it/~brunato/calendar/ and mail me.
• Elisa Cilia - E-mail cilia@disi.unitn.it

Program

• February 20, 2008 (Brunato)
  First lab assignment: building blocks for a simple web crawler.
• February 21, 2008 (Brunato)
  Crawling basics: parallel page fetching, DNS prefetching, two-level hashing for URLs.
• February 27, 2008 (Brunato)
  Robots exclusion file, robot traps, network structure of a crawler.
  Second lab assignment: creating a word index.
• February 28, 2008 (Cilia)
  Indexing: Information Retrieval system architecture, text operations, direct and inverted indexing, queries.
  Retrieval and ranking: recall and precision.
• March 05, 2008 (Cilia)
  Vector-Space Model, Relevance Feedback and Rocchio's Method.
• March 06, 2008 (Cilia)
  Probabilistic Relevance Feedback: Odds ratio and Bayesian Inference Networks.
  Advanced Issues: spamming, internal tag structure and hyperlinking, ranking of complex queries including phrases, approximate string matching, metasearch systems.
• March 12, 2008 (Brunato)
  Third lab assignment: write a short essay on the seminar by prof. Jehoshua "Shuki" Bruck.
• March 13, 2008 (Cilia, Brunato)
  Exercises.
  The Jaccard similarity index, straightforward calculation, probabilistic definition, approximation by means of a randomized algorithm based on permutations.
• March 19, 2008 (Brunato)
  The Jaccard similarity index, conclusion. Techniques for eliminating near duplicates.
  Introduction to PageRank.
• March 26, 2008 (Brunato)
  Fourth lab assignment: estimating page similarity.
  Exercises on PageRank.
• March 27, 2008 (Cilia)
  PageRank and HITS algorithms. Introduction to clustering, agglomerative clustering.
• April 2, 2008 (Cilia)
  Fifth lab assignment: computing PageRank.
  Exercises on PageRank and HITS.
• April 3, 2008 (Brunato)
  Clustering: k-means algorithm, Self Organizing Maps (SOMs).
  
• April 9, 2008 (Cilia)
  Introduction to geometric embedding approaches.
  Clustering and visualization via embeddings: Multidimensional scaling, FastMap, projection and subspaces.
• April 10, 2008 (Brunato)
  Singular value decomposition (see, for example, this tutorial ).
  Latent Semantic Indexing.
  Introduction to the Principal Component Analysis (see, for example, this tutorial).
• April 16, 2008 (Cilia)
  Sixth lab assignment: Applying the FastMap algorithm.
  Introduction to probabilistic approaches to clustering.
• April 17, 2008 (Brunato)
  Generative distributions for documents.
  Mixture models.
• April 23, 2008 (Brunato)
  Expectation Maximization algorithm.
  Lagrange multipliers.
• April 24, 2008 (Cilia)
  Multiple Cause Mixture Model (MCMM).
  Exercises on maximum likelihood and the least square method for parameter estimation.
• April 30, 2008 (Brunato)
  The MapReduce framework. [Reference: the Google paper]
  Final lab assignment
• May 07, 2008 (Cilia)
  Exercises on MapReduce.
• May 08, 2008 (Brunato)
  Introduction to collaborative filtering. Probabilistic models (Gibbs sampling) for the estimation of collaborative filtering parameters.
• May 14, 2008 (Brunato)
  Exercises on precision and recall.
  Exercises on Gibbs sampling for two-sided clustering.
• May 15, 2008 (Cilia)
  Exercises on relevance ranking, relevance feedback, LSI.
• May 21, 2008 (Brunato)
  Introduction to supervised learning: linearly separable problems, dealing with non-separability, the overfitting problem.
  [References: sections "Linear regression", "Using linear functions for classification", "Multi-layer perceptrons" from this bookchapter, entry "Overfitting" from Wikipedia.
• May 22, 2008 (Cilia)
  Exercises on LSI, HAC and HITS algorithm.

Page maintained by Mauro Brunato