Itinerary planning is a challenging task for users wishing to enjoy points of interest (POIs) in line with their preferences, the current context of use, and travel constraints. This article describes an approach to exploit linked open data (LOD) to perform a context-aware recommendation of personalized itineraries with related multimedia content. The recommendation process takes into account the user profile, the context of use, and the characteristics of the POIs extracted from LOD. The system, therefore, consists of six main modules that accomplish the following tasks: (i) the creation of the user profile according to her interests and preferences; (ii) the elicitation of the current context of use; (iii) the extraction and filtering of POIs from LOD through customized and dynamic queries; (iv) the itinerary construction to determine the first K itineraries that match the query; (v) their ranking through a score function that considers several factors, such as the POI popularity, the POI diversity in terms of their categories, the distance and the travel time of the itinerary, the user profile, and her physical and social context; (vi) the recommendation of multimedia and textual contents related to the itinerary suggested to the target user. The results of experimental tests performed on 50 real users show the benefits of the proposed recommender not only in terms of normalized discounted cumulative gain (nDCG), but also in terms of precision and beyond-accuracy metrics.

Fogli, A., Sansonetti, G. (2019). Exploiting semantics for context-aware itinerary recommendation. PERSONAL AND UBIQUITOUS COMPUTING, 23(2), 215-231 [10.1007/s00779-018-01189-7].

Exploiting semantics for context-aware itinerary recommendation

Sansonetti G.
2019-01-01

Abstract

Itinerary planning is a challenging task for users wishing to enjoy points of interest (POIs) in line with their preferences, the current context of use, and travel constraints. This article describes an approach to exploit linked open data (LOD) to perform a context-aware recommendation of personalized itineraries with related multimedia content. The recommendation process takes into account the user profile, the context of use, and the characteristics of the POIs extracted from LOD. The system, therefore, consists of six main modules that accomplish the following tasks: (i) the creation of the user profile according to her interests and preferences; (ii) the elicitation of the current context of use; (iii) the extraction and filtering of POIs from LOD through customized and dynamic queries; (iv) the itinerary construction to determine the first K itineraries that match the query; (v) their ranking through a score function that considers several factors, such as the POI popularity, the POI diversity in terms of their categories, the distance and the travel time of the itinerary, the user profile, and her physical and social context; (vi) the recommendation of multimedia and textual contents related to the itinerary suggested to the target user. The results of experimental tests performed on 50 real users show the benefits of the proposed recommender not only in terms of normalized discounted cumulative gain (nDCG), but also in terms of precision and beyond-accuracy metrics.
Fogli, A., Sansonetti, G. (2019). Exploiting semantics for context-aware itinerary recommendation. PERSONAL AND UBIQUITOUS COMPUTING, 23(2), 215-231 [10.1007/s00779-018-01189-7].
File in questo prodotto:
Non ci sono file associati a questo prodotto.

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/11590/373706
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 32
  • ???jsp.display-item.citation.isi??? 17
social impact