Personal recommendation based on a user's understanding

Abstract

Current web personalization mainly focuses on extracting web pages interesting to users.However, difficulty in understanding web contents often arises for some users mainly due to the lack of background knowledge on the contents. As a solution to this problem, this paper proposes a method of personalizing web search that produces search results suitable to the understanding level of a user. Fuzzy sets and membership functions are defined in order to keep adjusting the level of a user, which is updated based on the levels of web pages on which the user shows preferences. Extensive simulation experiments are conducted to tune the system parameters optimally and to show that the proposed system outperforms not only a general search engine but also a system that updates the understanding level of a user with the mean of the web page levels.

Fuzzy Tools in Recommender Systems: A Survey

“…Here it is worthy to note the development of tag-based user profiling methods for improving recommendations [9], where user profiles are built through a folksonomy-based approach that evaluates items according to the membership degrees to various attribute values, which are then used to compute the fuzzy user profile. Additionally, in the last few years further works on the use of fuzzy tools for modelling specific items' features in CBRS have been developed [2,5,14,69,96,99,104,126].…”

“…[99] Recommendation in online stores regarding marketing concepts Prec/Recall/F1 Non public Online stores Lee [69] Recommends items or web pages suitable to the users' understanding levels Prec/Recall/F1 Data generated by a simulator…”

Automated web content personalization in the automotive sector: The SEAT case

“…In addition to the interface and aesthetic design factors, there is also research on factors to be considered when designing the personalization algorithms. For example, some authors advocate for the inclusion of user feelings in the algorithms (Anand and Bharadwaj, 2013;Lee, 2012;Sahoo and Ratha, 2018) and others promote the use of user contextual factors such as timing of social communities (Hawalah and Fasli, 2014;Salonen and Karjaluoto, 2019;Paliouras, 2012).…”

“…• Content-based filtering: The user is seen as an individual and, thus, is suggested items or services which are similar to those bought or searched in the past, by matching the characteristics of the item or service with the characteristics of the user that are maintained in a the user profile (Adomavicius and Tuzhilin, 2005a;Hawalah and Fasli, 2015;Lee, 2012). The cornerstone of this method is the calculation of similarity between the items and the user profile information, and this is usually done by heuristics (such as cosine similarity), Bayesian classifiers and some machine learning techniques (Isinkaye et al, 2015;Lee, 2012;Pazzani, 1999). The main advantage of this method is that is based only on facts about the particular user and, ergo, are true (Kabassi, 2010).…”

“…And it is also possible to directly generate the recommendation in a system using a combination of the two methods (Adomavicius and Tuzhilin, 2005a). With it, the strengths of each of them are emphasized in order to improve the performance of a recommender system (Lee, 2012;Zhang et al, 2013).…”