First webpage time algorithm with mining domain name system data

Abstract

It is significant for an Internet Service Provider (ISP) or an Internet Content Provider (ICP) to obtain the user experience of the web service that is one of the most widely used Internet services. Mining data from the network systems to measure the web service has become one of the research hot spots recently. In this paper, we propose a novel method to obtain the user experience of the web service, the First Webpage Time (FWT) algorithm. The proposed FWT algorithm is performed by mining the Domain Name System (DNS) data that record the domain name resolution while users browse webpages. And then, it is analyzed in theory that there are a large number of FWTs to be averaged in order to ensure the precision of the final result. Furthermore, we also propose a new Key Quality Indicator (KQI) of measuring the web service, the FWT that is used to evaluate the loading time of textual contents of a webpage. The method of computing the FWT with DNS data is introduced and its correctness is discussed. At last, experiments based on ISP's DNS data are carried out to evaluate the proposed FWT algorithm.