Location: EU Supported » PASCAL - Pattern Analysis, Statistical Modelling and Computational Learning » Workshop on Subspace, Latent Structure and Feature Selection Techniques: Statistical and Optimisation Perspectives, Bohinj 2005

Latent Semantic Variable Models

author: Thomas Hofmann, Google
published: Feb. 25, 2007,   recorded: February 2005,   views: 2959
Categories
You might be experiencing some problems with Your Video player.

Slides

Slides
0:01 Latent Semantic Variable Models
0:43 Introduction Information Retrieval & Latent Semantic Indexing Probabilistic Latent Semantic Indexing Semantic Features for Text Categorizisation Probabilistic HITS Collaborative Filtering Concl
1:19 Latent Structure
2:23 Matrix Decomposition
3:13 Introduction Information Retrieval & Latent Semantic Indexing Probabilistic Latent Semantic Indexing Semantic Features for Text Categorizisation Probabilistic HITS Collaborative Filtering Concl
3:23 Searching & Finding
4:14 Ad Hoc Retrieval
4:55 Document-Term Matrix
6:47 A 100 Millionths of a Typical Document-Term Matrix
7:45 Robust Information Retrieval — Beyond Keyword-based Search
9:48 Challenges
10:41 Latent Semantic Analysis
12:18 Singular Value Decomposition
13:26 Low-rank Approximation
14:35 LSA Decomposition
15:21 Latent Semantic Analysis
16:31 Introduction Information Retrieval & Latent Semantic Indexing Probabilistic Latent Semantic Indexing Semantic Features for Text Categorizisation Probabilistic HITS Collaborative Filtering Concl
16:39 Search as Statistical Inference
19:00 Language Model Paradigm in IR
20:10 Language Model Paradigm
20:46 Language Model Paradigm
21:13 Naive Approach
21:51 Estimation Problem
22:38 Probabilistic Latent Semantic Analysis
24:13 pLSA – Latent Variable Model
25:25 pLSA: Matrix Decomposition
27:19 pLSA: Graphical Model
28:37 pLSA via Likelihood Maximization
29:26 Expectation Maximization Algorithm
31:22 EM Algorithm: Derivation
31:41 Tempered EM Algorithm
35:05 Example (1)
37:32 Example (2)
38:13 Experimental Evaluation
39:41 Live Implementation
43:08 Latent Dirichlet Allocation
46:17 Introduction Information Retrieval & Latent Semantic Indexing Probabilistic Latent Semantic Indexing Semantic Features for Text Categorizisation Probabilistic HITS Collaborative Filtering Concl
46:32 Concept-based Text Categorization
47:32 Terms & Concepts as Features
48:36 Improvements on Reuters-21578
49:32 Improvements on OHSUMED87
50:59 Literature & Related Work
52:00 Introduction Information Retrieval & Latent Semantic Indexing Probabilistic Latent Semantic Indexing Semantic Features for Text Categorizisation Probabilistic HITS Collaborative Filtering Concl
52:01 Probabilistic HITS
53:40 Finding Latent Web Communities
54:06 Decomposing the Web Graph
55:01 Linking Hyperlinks and Content
56:06 Example: Ulysses
57:38 Literature & Related Work
57:40 Introduction Information Retrieval & Latent Semantic Indexing Probabilistic Latent Semantic Indexing Semantic Features for Text Categorizisation Probabilistic HITS Collaborative Filtering Concl
57:51 Predictions & Recommendations

Related content

Report a problem or upload files

If you have found a problem with this lecture or would like to send us extra material, articles, exercises, etc., please use our ticket system to describe your request and upload the data.
Enter your e-mail into the 'Cc' field, and we will keep you updated with your request's status.
Lecture popularity: You need to login to cast your vote.
 
    Delicious Bibliography

Description

In the context of information retrieval and natural language processing, latent variable models are quite useful in modeling and discovering hidden structure that often leads to "semantic" data representations. This talk will provide an overview of the most popular approaches and discuss the range of possible applications for such models, including language modeling, ad hoc retrieval, text categorization and collaborative filtering.

See Also:

Launch Windows Media PlayerLaunch in a standalone WM Player

WMedia PlayerSwitch to Windows Media Player

Download slides icon Download slides: slsfs05_hofmann_lsvm_01.ppt (3.8 MB)

Help icon Streaming Video Help

WebLink icon Windows Media Player Firefox Plugin - Download

Link this page

Would you like to put a link to this lecture on your homepage?
Go ahead! Copy the HTML snippet !

Reviews and comments:

Comment1 Mike, March 12, 2008 at 1:18 p.m.:

Interesting talk, too bad the last part about recommenders is missing.

Comment2 R., May 23, 2008 at 8:18 p.m.:

On the first EM slide, should the bottom right equation be P(d|z; pi) instead of P(z|d; pi)?

Comment3 Andrew Polar, October 6, 2011 at 7:24 a.m.:

Thomas Hufmann over-complicates real nature of his algorithm on purpose. It can be perfectly clear explained in linear algebra notations. Linear algebra excludes obscurity. Here is the link where everything is explained on 1 page:
http://semanticsearchart.com/research...

Write your own review or comment:

make sure you have javascript enabled or clear this field: