Terry+-+Annotated+Bibliography

This article determines how websites these days make use of recommender systems to allow users to get access to more relative information that they might require. In particular, it examines the different web content mining techniques used to create a recommender system for a website. Recommender systems need to be constantly updated. To deliver more accurate information for users’ requests, the authors proposed a hybrid recommender model by combining results of different recommender techniques. Collaborative Filtering (CF), a highly popular and power technique used to build recommendation system was also discussed. The paper concluded that this hybrid recommender model is still developing and could pave the way for future recommender systems.
 * Göksedef, Murat, and Şule Gündüz-Öğüdücü . "Combination of Web page recommender systems." //Expert Systems with Applications//. 37.4 (2010): 2911-2922. Print.**

**Ant Ozok, A., Quyin Fan, and Anthony F. Norcio. "Design guidelines for effective recommender system interfaces based on a usability criteria conceptual model: results from a college student population ." //Behaviour & Information Technology//. 29.1 (2010): 57-83. Print.** This article evaluates the current recommender system used in most web pages today. Specifically, the article examines the user interface design of these systems and identifies that user preference issues are key areas to be looked at. This is to determine if the system is presented correctly to users. Often, a poorly designed user interface would lead to an unpleasant customer shopping experience and would deter them from returning. The article is of good relevance as it uses the college student population as the main target group for their study. These are the people that are most susceptible to e-commerce. The authors concluded that transparency and sufficiency issues are most essential to building an effective recommender system.

This article reviews that improvements need to be made to recommender systems to provide a more accurate and satisfying recommendation result to users. The authors recognize that majority of the recommendation systems focus on algorithm factors to improve its reliability. Unfortunately, they all face a common problem which is the use of information data that they collect under a similar environment. The article suggests that by applying a multidimensional model, together with online analytical processing, it would gather data of customer’s perceptions and provide a more satisfactory recommendation result.
 * Weng, Sung-Shun, Binshan Lin, and Wei-Tien Chen. " Using contextual information and multidimensional approach for recommendation." //Expert Systems with Applications//. 36.2 (2009): 1268-1279. Print.**

The article seeks to explain two approaches which will improve the search result quality of users browsing using tagging methods. The article identifies that the strategies used in recommendation systems today, content –based and collaborative filtering, do not produce accurate search results. The authors suggest that content-based strategy used in recommendation systems can produce better results if it applies a “query expansion”, which is basically an extension to a user’s questions and bombarding him with more than enough suggestions. Another approach discussed is to improve the profile of a user by recognizing his requirements. In this context, collaborative filtering will bring in more appropriate resources for him.
 * De Meo, Pasquale, Giovanni Quattrone, and Domenico Ursino. "A query expansion and user profile enrichment approach to improve the performance of recommender systems operating on a folksonomy." //User Modeling and User - Adapted Interaction [Dordrecht]//. 20.1 (2010): 41-86. Print.**

The paper discusses primarily about recommendation system that is used in a work environment. The article focuses on a system model that is suited for an organization that employs a hierarchical structure; whereby members perform individual specialized tasks and work together in teams often. The authors have also identified two key issues in the recommendation system to be looked at, namely classifying which domains are suitable and the amount of information one has. They argue that despite the current recommendation model and methods in place, there are certain boundaries yet to be discussed. Examples include recognizing the “expert level” of a new member and conducting experiments in all possible work environments.
 * Zhen, Lu, George Q. Huang, and Zuhua Jiang. "Recommender system based on workflow." //Decision Support Systems//. (2009): 237-245. Print.**