Abstract: In recent years there have been various attempts and studies that are eager to serendipity in Computer Science. Authors such as Campos and Dias have tried to model the serendipity in order to get serendipitous behaviors in Information Retrieval (IR). There have been attempts to introduce serendipity in Recommender Systems (RS), although the latter proposals have led efforts using metrics for measure Serendipity in those RS, rather than to emulate Serendipitous behaviors in recommender systems. However, so far there haven´t been succeeded in designing a model which can be applied to different Web browsing. The main problem we have detected analyzing the proposals in this field is that the solutions provided do not take into account the two aspects of the concept of serendipity. Do not consider that in addition to the accidental discovery of information that is not sought for, is also required some characteristics in the user like sagacity, perception, flexible thinking or intensive preparation. If we could develop a model that support any search engine or search tool, we would facilitate an incredible advantage to the user by offering specially information that the user is not focused upon. The aim of this paper is to propose a computational model that supports Serendipity and induce and facilitate serendipity through the use of a special-purpose designed system.

Keywords: Supporting serendipity, : Designing Serendipity Serendipia, Data Minig, Artificial intelligence, SOM, Models of Computation

ACM Classification Keywords: H.2.8 Data Mining, F.1.1 Models of Computation, I.2 Artificial intelligence

Link:

COMPUTATIONAL MODEL FOR SERENDIPITY

A. Anguera, M.A. Diaz, A. Gutierrez

http://foibg.com/ijitk/ijitk-vol05/ijitk05-1-p08.pdf