Information and Recommender Systems

Elsa Negro has a PhD in computer science and is professor at Paris-Dauphine University, France. She is particularly interested in decision-making information systems, recommendation systems, intelligent digital cities and crisis management.

• Introduction viiChapter 1. A Few Important Details Before We Begin 11.1. Information systems 11.2. Decision support systems 21.3. Recommender systems 31.4. Comparisons 41.5. Recommendation versus personalization 51.5.1. Recommendation 51.5.2. Personalization 6Chapter 2. Recommender Systems 72.1. Introduction 82.2. Classification of recommender systems 92.2.1. Classification by score estimation method 92.2.2. Classification by data exploitation 102.2.3. Classification by objective 112.3. User profiles 112.4. Data mining 122.5. Content-based approaches 142.6. Collaborative filtering approaches 172.7. Knowledge-based approaches 202.8. Hybrid approaches 232.9. Other approaches 25Chapter 3. Key Concepts, Useful Measures and Techniques 293.1. Vector space model 313.2. Similarity measures 313.2.1. Cosine similarity 313.2.2. Pearson correlation coefficient 323.2.3. Euclidean distance 333.2.4. Dice index 333.3. Dimensionality reduction 343.3.1. Principal component analysis 343.3.2. Singular value decomposition 353.3.3. Latent semantic analysis 363.4. Classification/clustering 363.4.1. Classification 363.4.2. Clustering 373.5. Other techniques 393.5.1. Term frequency-inverse document frequency (TF-IDF) 393.5.2. Association rules 403.6. Comparisons 41Chapter 4. Practical Implementations 434.1. Commercial applications 434.1.1. Amazon.com 434.1.2. Netflix 454.2. Databases 464.3. Collaborative environments 484.4. Smart cities 494.5. Early warning systems 54Chapter 5. Evaluating the Quality of Recommender Systems 575.1. Data sets, sparsity and errors 575.2. Measures 595.2.1. Accuracy 595.2.2. Other measures 63Conclusion 65Bibliography 67Index 77