The World Wide Web has become the most frequently used and highest volume computer transmission protocol ever devised. This growth can be observed using the NSFNET performance statistics that have been collected and maintained by the Merit Network since the early days of the NSFNET project[Merit Network, Inc.1995]. The use of the WWW, measured in the bytes transferred, has show exponential growth from November 1992, the time such statistics started to be collected, to April 30, 1995, when the NSFNET Backbone Service was discontinued and replaced by a new network architecture. The volume of network traffic in terabytes (i.e. units of 1,000,000,000,000 bytes) on the NSFNet for 5 different protocols (WWW, FTP, Gopher, Mail, and Usenet News) are shown in Figure 1.
late in 1994 these statistics started to decrease as NSFNET traffic began to migrate to the new internet architecture. The November 1994 statistics were the last to reflect the ``pure'' NSFNet traffic in its entirety. However, a sudden spike in all forms of network traffic may be seen in early 1995 as commercial online vendors like America On-Line and Compuserve began to offer internet access to their customers. However, looking at the percentage of each type of internet traffic, as shown in Figure 2, the exponential growth trend of the WWW is obvious. We can only speculate as to the current rate of growth for the internet and WWW traffic, because the new network architecture no longer allows (or the carriers don't supply) this information. However, if we extrapolate from the growth curve observed in Figure 2 we can assume that the WWW is now accounting for over 60% of internet traffic.
One thing that can be known is how much information is currently available in WWW documents. Between July and November of 1995 a research group at Berkeley [Aoki et al.1996] examined documents collected by the ``webcrawler'' program used to generate the indexes for the Inktomi network search engine. The first complete ``crawl'' of the WWW carried out between July and October yielded 1.3 million documents. In November 1995 a second crawl returned over 2.6 million documents. Although this is only a snapshot of the WWW, the fact that it effectively doubled in size in a single month seems to provide some confirmation that the growth in the access traffic has been paralleled by an exponential growth in content as well.