| 01 | Boolean retrieval | Inverted index, postings intersection, query optimization | ch1 |
| 02 | The term vocabulary and postings lists | Phrase queries "Stanford University" and proximity queries Gates near Microsoft → the index must store word positions too | ch2 |
| 03 | Dictionaries and tolerant retrieval | k-gram index: bo → aboard, about, boardroom, border. Find words even when misspelled | ch3 |
| 04 | Index construction | Split a large collection across many machines (parser → inverter, map·reduce) | ch4 |
| 05 | Index compression | Zipf's law: log frequency vs log rank is a straight line | ch5 |
| 06 | Scoring, term weighting, the vector space model | Boolean only includes/excludes → rank by query-document proximity. Needs a function that turns occurrence counts into scores | ch6 |
| 07 | Computing scores in a complete search system | The whole structure: parser, index, spelling correction, scoring, ranking | ch7 |
| 08 | Evaluation, dynamic summaries | | Out of scope |
| 09 | Relevance feedback, query expansion | | Out of scope |
| 12 | Language models | Unigram model: multiply word probabilities to get a string's probability | Out of scope |
| 13 | Text classification, Naive Bayes | Automatically assign documents to predefined classes (language, adult content, region) | Out of scope |
| 14 | Vector classification | | Out of scope |
| 15 | SVM | | Out of scope |
| 16 | Flat clustering | | Out of scope |
| 17 | Hierarchical clustering | | Out of scope |
| 18 | Latent semantic indexing (LSI) | | Out of scope |
| 19 | The web and its challenges | Diverse documents and users, using link analysis and user data | Out of scope |
| 21 | Link analysis, PageRank | | Out of scope |
