Web Information Extraction and Retrieval

Content of the course:
This course will cover the following topics:

• Information Retrieval and Web Search:
  Basic Concepts of Information Retrieval
  Information Retrieval Models
  Relevance Feedback
  Evaluation Measures
  Text and Web Page Pre-Processing
  Inverted Index and Its Compression
  Latent Semantic Indexing
  Web Search
  Meta-Search: Combining Multiple Rankings

• Web Crawling:
  A Basic Crawler Algorithm
  Implementation Issues
  Universal Crawlers
  Focused Crawlers
  Topical Crawlers

• Structured Data Extraction:
  Wrapper Induction
  Instance-Based Wrapper Learning
  Automatic Wrapper Generation
  String Matching and Tree Matching
  Multiple Alignment
  Building DOM Trees
  Extraction Based on a Single List Page or Multiple Pages

• Information Integration:
  Schema-Level Matching
  Domain and Instance-Level Matching
  Combining Similarities
  1:m Match
  Integration of Web Query Interfaces
  Constructing a Unified Global Query Interface

• Opinion Mining and Sentiment Analysis:
  Document Sentiment Classification
  Sentence Subjectivity and Sentiment Classification
  Opinion Lexicon Expansion
  Aspect-Based Opinion Mining
  Opinion Search and Retrieval

• Bing Liu, Web Data Mining: Exploring Hyperlinks, Contents, and Usage Data (Data-Centric Systems and Applications, Springer, August 2013
• Ricardo Baeza-Yates , Berthier Ribeiro-Neto: Modern Information Retrieval: The Concepts and Technology behind Search, 2nd Edition, ACM Press Books, 2010

The main objective of this course is to teach students about how to develop programs for web search (including surface web and deep web search) and for extraction of structural data from both, static and dynamic web pages. Beside basic concepts of the web search and retrieval, students will learn about relevant techniques and approaches. After the course, if successful, students will be able to develop programs for automatic web search and structured data extraction from web pages (including search and extraction from on-line social media).

After successful completion of the module, students will be able to:

• summarize the most important approaches and techniques for searching and extracting data from the web
• to select approaches and techniques that are most suitable for individual problems in web information extraction and retrieval.
• to develop applications for data acquisition and analysis,
• to construct new algorithms for web data search and extraction,
• to explain behavior and time complexity of specific web search algorithms,
• to integrate and employ different open-source solutions from the field.

Lectures, seminars, homeworks, oral presentations, project work.

Continuing (homework, midterm exams, project work)
Final (written and oral exam)
grading: 5 (fail), 6-10 (pass) (according to the Statute of UL)

Pet najpomembnejših del:
ŠUBELJ, Lovro, BAJEC, Marko. Group detection in complex networks : an algorithm and comparison of the state of the art. Physica. A, 2014
ŽITNIK, Slavko, ŠUBELJ, Lovro, LAVBIČ, Dejan, VASILECAS, Olegas, BAJEC, Marko. General context-aware data matching and merging framework. Informatica, 2013
LAVBIČ, Dejan, BAJEC, Marko. Employing semantic web technologies in financial instruments trading : Dejan Lavbič and Marko Bajec. International journal of new computer architectures and their applications, 2012
ŠUBELJ, Lovro, FURLAN, Štefan, BAJEC, Marko. An expert system for detecting automobile insurance fraud using social network analysis. Expert systems with applications, 2011
ŠUBELJ, Lovro, JELENC, David, ZUPANČIČ, Eva, LAVBIČ, Dejan, TRČEK, Denis, KRISPER, Marjan, BAJEC, Marko. Merging data sources based on semantics, contexts and trust. The IPSI BgD transactions on internet research, 2011
Celotna bibliografija je dostopna na SICRISu:
http://sicris.izum.si/search/rsr.aspx?lang=slv&,id=9270.