Web Mining

Basic Concepts of Information Retrieval (IR). IR Methods – Boolean Model, Vector Space Model and Statistical Language Model. Relevance Feedback. Evaluation Measures. Text Preprocessing- Stopword Removal, Stemming. Web Page Preprocessing-Duplicate Detection. Inverted Index and Its Compression – Inverted Index, Search using Inverted Index, Index Construction, Index Compression. Meta-Search and Combining Multiple Rankings. Web Spamming- Content spamming, Link spamming.

Link Analysis – Social Network Analysis. Co-Citation and Bibliographic Coupling. Page Rank-Algorithm, Strengths and Weaknesses, Timed pageRank. HITS-Algorithm, Relationships with Co-Citation and Bibliographic Coupling, Strengths and Weaknesses of HITS. Community Discovery-Problem Definition, Bipartite Core Communities, Maximum Flow Communities, Email Communities Based on Betweenness, Overlapping Communities of Named Entities.

Web Crawling – A Basic Crawler Algorithm-Breadth First Crawlers, Preferential Crawlers. Implementation Issues- Fetching, Parsing, Stopword Removal and stemming, Link Extraction and Canonicalization, Spider Traps, Page Repository, Concurrency. Universal Crawlers, Focused Crawlers, Topical Crawlers, Crawler Ethics and Conflicts.

Sentiment Classification –Classification based on Sentiment Phrases, Classification Using Text Classification Methods, Classification Using a Score Function. Feature based Opinion Mining and Summarization – Problem Definition, Object feature extraction, Feature Extraction from Pros and Cons of Format1, Feature Extraction from Reviews of Format 2 and 3, Opinion Orientation Classification. Comparative Sentence and Relation Mining, Opinion Search and Opinion Spam.

Data Collection and Preprocessing- Sources and Types of Data, Key Elements of Web usage Data Preprocessing. Data Modeling for Web Usage Mining. Discovery and Analysis of Web usage Patterns -Session and Visitor Analysis, Cluster Analysis and Visitor Segmentation, Association and Correlation Analysis, Analysis of Sequential and Navigation Patterns.