We propose an approach based on Swarm Intelligence -- more specifically on Ant Colony Optimization (ACO)-- to improve search engines’ performance and reduce information overload by exploiting collective users’ behavior. We designed and developed three different algorithms that employ an ACO-inspired strategy to provide implicit collaborative-seeking features in real time to search engines. The three different algorithms -- NaïveRank, RandomRank, and SessionRank -- leverage on different principles of ACO in order to exploit users’ interactions and provide them with more relevant results. We designed an evaluation experiment employing two widely used standard datasets of query-click logs issued to two major Web search engines. The results demonstrated how each algorithm is suitable to be employed in ranking results of different types of queries depending on users’ intent.

An ant-colony based approach for real-time implicit collaborative information seeking / Malizia, Alessio*; Olsen, Kai A.; Turchi, Tommaso; Crescenzi, Pierluigi. - In: INFORMATION PROCESSING & MANAGEMENT. - ISSN 0306-4573. - STAMPA. - 53:(2017), pp. 608-623. [10.1016/j.ipm.2016.12.005]

An ant-colony based approach for real-time implicit collaborative information seeking

Crescenzi, Pierluigi
2017

Abstract

We propose an approach based on Swarm Intelligence -- more specifically on Ant Colony Optimization (ACO)-- to improve search engines’ performance and reduce information overload by exploiting collective users’ behavior. We designed and developed three different algorithms that employ an ACO-inspired strategy to provide implicit collaborative-seeking features in real time to search engines. The three different algorithms -- NaïveRank, RandomRank, and SessionRank -- leverage on different principles of ACO in order to exploit users’ interactions and provide them with more relevant results. We designed an evaluation experiment employing two widely used standard datasets of query-click logs issued to two major Web search engines. The results demonstrated how each algorithm is suitable to be employed in ranking results of different types of queries depending on users’ intent.
2017
53
608
623
Malizia, Alessio*; Olsen, Kai A.; Turchi, Tommaso; Crescenzi, Pierluigi
File in questo prodotto:
File Dimensione Formato  
IPM-2017.pdf

Accesso chiuso

Tipologia: Pdf editoriale (Version of record)
Licenza: Tutti i diritti riservati
Dimensione 821.93 kB
Formato Adobe PDF
821.93 kB Adobe PDF   Richiedi una copia
Malizia et al - Manuscript.pdf

accesso aperto

Tipologia: Altro
Licenza: Tutti i diritti riservati
Dimensione 361.39 kB
Formato Adobe PDF
361.39 kB Adobe PDF

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

Utilizza questo identificatore per citare o creare un link a questa risorsa: https://hdl.handle.net/2158/1113524
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 10
  • ???jsp.display-item.citation.isi??? 4
social impact