Digital Libraries and Archives: 7th Italian Research by Claudio Taranto, Nicola Di Mauro, Floriana Esposito (auth.),

By Claudio Taranto, Nicola Di Mauro, Floriana Esposito (auth.), Maristella Agosti, Floriana Esposito, Carlo Meghini, Nicola Orio (eds.)

This ebook constitutes the completely refereed court cases of the seventh Italian examine convention on electronic Libraries held in Pisa, Italy, in January 2011. The 20 revised complete papers provided have been conscientiously reviewed and canopy subject matters of curiosity corresponding to process interoperability and knowledge integration; formal and methodological foundations of electronic libraries; semantic internet and associated information for electronic libraries; multilingual details entry; electronic library infrastructures; metadata production and administration; se's for electronic library platforms; assessment and log info; dealing with audio/visual and non-traditional gadgets; consumer interfaces and visualization; electronic library quality.

Show description

Read or Download Digital Libraries and Archives: 7th Italian Research Conference, IRCDL 2011, Pisa, Italy, January 20-21, 2011. Revised Papers PDF

Similar italian books

Additional info for Digital Libraries and Archives: 7th Italian Research Conference, IRCDL 2011, Pisa, Italy, January 20-21, 2011. Revised Papers

Example text

5 Conclusion and Future Work In this work we introduced a content-based paper recommender system which produces rich user profiles and resource descriptions by extracting keyphrases from scientific articles. The system is intended to work in combination with other modules of an hosting framework, currently under development within Pirates, a larger project aimed at innovate within a social/semantic approach the tools for access, classification, filtering, retrieval, and extraction of Web information.

38–48, 2011. c Springer-Verlag Berlin Heidelberg 2011 Improving User Stereotypes through Machine Learning Techniques 39 A way to overcome such a limitation could be represented by the exploitation of models built on behaviours instead of explicitly declared user interests. : university, hospital, business office). A challenge, in this respect, is the inability of the systems to meet individual user expectations at run-time. A step in this direction could be done by exploiting machine learning techniques, but this requires approaches that are specific to the task.

Differently, the approach presented in this paper proposes to exploit the knowledge that is extracted by the analysis of log interaction data, without requiring an explicit feedback from the user, in a cascaded unsupervised and supervised machine learning techniques to improve the set of user stereotypes. 3 The Framework The general framework we propose, depicted in Figure 1, is made up of a module aimed at creating a set of user classes, followed by a module devoted to generate, on these classes, a set of rules to be used for classifying the behaviour of new and unseen users.

Download PDF sample

Rated 4.70 of 5 – based on 12 votes