Un servizio di web caching e prefetching guidato da profili d'utenza (APWebCache)
Autore
Giuseppe Patelli - Politecnico di Milano - [1997-98]
Documenti
Abstract
As the number of World Wide Web (WWW) users grows, the perceived latency of the WWW is becoming an important issue. The retrieval time of WWW documents depends on several factors such as network bandwidth, propagation delay and host performance.
We focused on the mechanisms provided by caching and prefetching technologies in order to reduce WWW latency. Caching technology aims at reducing network traffic and latency while prefetching aims at reducing propagation delay by separating request time and utilization time.
We designed and developed a prototype, named APWebCache which stands for Active and Personalised Web Cache, able to deploy these technologies integrated with a recommendation system ordinarily used in web marketing. Further, personalization and collaboration techniques were added.
The prototype offers a caching service based on a traditional proxy and an asynchronous prefetching service driven by user profiles and a collaborative filtering engine.
We focused on the mechanisms provided by caching and prefetching technologies in order to reduce WWW latency. Caching technology aims at reducing network traffic and latency while prefetching aims at reducing propagation delay by separating request time and utilization time.
We designed and developed a prototype, named APWebCache which stands for Active and Personalised Web Cache, able to deploy these technologies integrated with a recommendation system ordinarily used in web marketing. Further, personalization and collaboration techniques were added.
The prototype offers a caching service based on a traditional proxy and an asynchronous prefetching service driven by user profiles and a collaborative filtering engine.
Questa tesi è correlata alla categoria