Machine learning for the semantic web

Published on 2011-09-055603 Views

SSSC 2011 - Berkeley

Watch other parts:

Presentation

Document Classification into large taxonomies00:00

How to Represent Text? ...from Characters to Logic00:00

DMoz (Open Directory Project)02:20:56

Some Initial thoughts17:43:34

Classification of a query into DMoz40:08:38

Quick example: Why representation matters?46:47:34

Classification of a document into DMoz66:36:08

Visual & Contextual Search81:19:19

Contextualized search85:06:47

Example: searching 167:53:03

Context sensitive search with168:07:34

News reporting bias168:16:57

News Reporting Bias example173:54:39

Experimental setup181:54:54

Prediction of news source190:22:40

Detecting News Reporting194:20:51

Easy decision problems require simple data representation210:08:40

Harder decision problems require better data representation - 1210:28:27

Harder decision problems require better data representation - 2210:50:19

Harder decision problems require better data representation - 3211:15:17

How we represent Text?211:29:47

How we process data?211:41:01

News Visualization212:05:14

Topic landscape of the query “Clinton” from Reuters news 1996-1997213:23:21

What we do with data?263:31:38

Visualization of social relationships between “Clinton”263:39:24

Topic Trends Tracking of the documents including “Clinton”272:39:29

Key paradigms287:36:05

WW2 query “Pearl Harbor” into NYTimes archive305:09:01

WW2 query “Belgrade” into NYTimes archive315:23:36

How different research areas approach text?348:18:09

WW2 query “Normandy” into NYTimes archive349:54:40

Levels of text representations - 9356:26:43

Language model level360:59:13

Context aware auto-complete373:07:18

Context-aware prediction for document authoring - 1374:52:00

Context-aware prediction for document authoring - 2375:48:45

How do we represent text?391:07:21

Context-aware prediction for document authoring - 3392:02:20

Levels of text representations - 2392:58:08

Levels of text representations - 10404:25:21

Full-parsing level404:53:24

Text Enrichment412:59:57

Text enrichment with http:// Enrycher.ijs.si413:30:08

Levels of text representations - 1420:57:38

Levels of text representations - 3487:58:57

Character level representation489:05:33

Good and bad sides of character level representation (n-grams)502:28:11

Language identification508:42:12

Graph559:39:03

Knowledge based summarization560:05:02

Summarization via semantic graphs560:27:55

Detailed Summarization560:37:18

Example of automatic summary560:42:40

Character level normalization585:58:30

Levels of text representations - 4586:30:42

Word level602:24:54

Key semantic word Properties606:22:58

Levels of text representations - 12619:40:57

Collaborative tagging630:13:27

Example: flickr.com tagging630:21:46

Example: del.icio.us tagging630:34:12

Stop-words631:21:06

Levels of text representations - 14635:28:45

Stemming and lemmatization638:58:33

Template / frames level639:38:02

Examples of simple templates644:13:16

Stemming659:02:04

Levels of text representations - 5669:29:32

Phrase level670:47:45

Google n-gram corpus686:32:09

Examples of Google n-grams709:01:38

Levels of text representations - 6717:03:57

Part-of-Speech718:28:12

Part-of-Speech Tags718:51:03

Part-of-Speech examples719:14:59

Levels of text representations - 7732:10:49

Taxonomies/thesaurus level733:05:40

WordNet – database of lexical734:40:08

WordNet relations744:19:58

Levels of text representations - 8768:52:32

Vector-space model level773:35:47

Cyc’s front-end: “Cyc Analytic Environment” – querying 784:17:03

Bag-of-Words document representation786:30:32

Bag-of-Words Words799:28:09

Example document and its vector representation819:05:18

Cyc’s front-end: “Cyc Analytic Environment” – justification 825:49:47

Similarity between BoW845:40:14

Document Categorization Task876:43:20

Algorithms for learning document classifiers888:40:06

Example learning algorithm: Perceptron900:26:35

Measuring success – Model quality estimation914:29:55

Document Clustering Task915:21:47

K-Means clustering algorithm923:30:08

Example of hierarchical clustering (bisecting k-means)926:59:16

Latent Semantic Indexing928:24:33

LSI Example928:43:57

Further references ...963:48:48

References to some Text-Mining books964:03:21

Books on Semantic Technologies968:54:54

References to the main conferences970:44:07

Videos on Text and Semantic Technologies976:03:01