Collaborative Filtering by Personality Diagnosis: A Hybrid Memory- and\n Model-Based Approach

2013/01/16 by David M. Pennock, Eric Horvitz, Pennock, David M. +5 · 3 citations
Business, Management and Accounting · Computer Science · Psychology · #Customer churn and segmentation #Data Mining Algorithms and Applications #FOS: Computer and information sciences #Information Retrieval (cs.IR) #Personality Traits and Psychology #Recommender Systems and Techniques

paper · pdf · doi:10.48550/arxiv.1301.3885

openalex publication_date 2013/01/16 · openalex created_date 2025/10/24 · openalex updated_date 2026/07/28

Abstract

The growth of Internet commerce has stimulated the use of collaborative\nfiltering (CF) algorithms as recommender systems. Such systems leverage\nknowledge about the known preferences of multiple users to recommend items of\ninterest to other users. CF methods have been harnessed to make recommendations\nabout such items as web pages, movies, books, and toys. Researchers have\nproposed and evaluated many approaches for generating recommendations. We\ndescribe and evaluate a new method called emphpersonality diagnosis (PD).\nGiven a user's preferences for some items, we compute the probability that he\nor she is of the same "personality type" as other users, and, in turn, the\nprobability that he or she will like new items. PD retains some of the\nadvantages of traditional similarity-weighting techniques in that all data is\nbrought to bear on each prediction and new data can be added easily and\nincrementally. Additionally, PD has a meaningful probabilistic interpretation,\nwhich may be leveraged to justify, explain, and augment results. We report\nempirical results on the EachMovie database of movie ratings, and on user\nprofile data collected from the CiteSeer digital library of Computer Science\nresearch papers. The probabilistic framework naturally supports a variety of\ndescriptive measurements - in particular, we consider the applicability of a\nvalue of information (VOI) computation.\n

Cited by

• Probabilistic Models for Unified Collaborative and Content-Based Recommendation in Sparse-Data Environments
• Symmetric Collaborative Filtering Using the Noisy Sensor Model
• Evaluating Recommendation Algorithms by Graph Analysis
• A Bayesian Approach toward Active Learning for Collaborative Filtering
• Evaluating collaborative filtering recommender systems
• Real-World Recommender Systems for Academia: The Pain and Gain in\n Building, Operating, and Researching them [Long Version]
• A Hybrid Approach to Web Service Recommendation Based on QoS-Aware Rating and Ranking
• CryptoRec: Privacy-preserving Recommendation as a Service
• Slope One Predictors for Online Rating-Based Collaborative Filtering
• Elicitation of Factored Utilities

Related

• Trust-Based Collaborative Filtering: Tackling the Cold Start Problem Using Regular Equivalence
• A Latent Source Model for Online Collaborative Filtering
• Symmetric Collaborative Filtering Using the Noisy Sensor Model
• Explainable Recommendations via Attentive Multi-Persona Collaborative\n Filtering
• Learning Complex Users' Preferences for Recommender Systems