Recommendation Systems Comparison Sample Clauses

Recommendation Systems Comparison. Due to the numerous possible approaches regarding a recommendation system, it is considered necessary to create a comparison section, in order to clearly represent the capabilities of each type of recommendation and how similar it is with other recommender systems. The tables in this section attempt to prepare the definitions for the implementation phase, describing a more practical representation of the algorithms. Additionally, this section describes how methods and algorithms described in Section 3.2 relate to the recommender systems. This is described throughout Table 2. The table is separate in rows, referring to the types of recommendation systems. Additionally, the table columns compare the two ways that the recommender systems can be implemented. Thus, consequently its cells are filled with information regarding the methods and algorithms that are used in each approach. Heuristic-based Model-based Content-based Commonly used techniques: • TF-IDF (information retrieval) • Clustering Representative research examples: • Lang 1995 • Balabovic & Shoham 1997 • Pazzani & Billsus 1997 Commonly used techniques: • Bayesian classifiers • Clustering • Decision trees • Artificial neural networks Representative research examples: • Xxxxxxx & Billsus 1997 • Xxxxxx et al. 1998 • Xxxxxx & Xxx 1999 • Billsus & Xxxxxxx 1999, 2000 • Xxxxx et al. 2002 Collaborative Commonly used techniques: • Nearest neighbour (cosine, correlation) • Clustering • Graph theory Representative research examples: • Xxxxxxx et al. 1994 • Hill et al. 1995 • Xxxxxxxxxx & Xxxx 1995 • Xxxxxx et al. 1998 • Xxxxxxxx & Xxx 1998 Commonly used techniques: • Bayesian networks • Clustering • Artificial neural networks • Linear regression • Probabilistic models Representative research examples: • Billsus & Xxxxxxx 1999, 2000 • Xxxxxx et al. 1998 • Xxxxx & Xxxxxx 1998 • Chien & Xxxxxx 1999 • Xxxxxxxx et al. 1999 • Xxxxxxx & Xxxxx 1999 • Xxxxxxx & Xxxxxxx 1999 • Sarwar et al. 2001 • Getoor & Sahami 1999 • Xxxxxxx & Xxxxxxx 1999 • Xxxxxxxx et al. 2001 • Xxxxx et al. 2001 • Pavlov & Xxxxxxx 2002 • Xxxxx et al. 2002 • Xx et al. 2002, 2004 • Xxxxxxx 2003, 2004 • Xxxxxx 2003 • Si & Jin 2003 Hybrid Combining content-based and collaborative components using: • Linear combination of predicted ratings • Various voting schemes • Incorporating one component as a part of the heuristic for the other Representative research examples: • Xxxxxxxxxxx & Shoham 1997 • Xxxxxxxx et al. 1999 • Good et al. 1999 • Pazzani 1999 ...