1. Access Analytics

  1. Monthly Time Series / Yearly Time Series
  2. Spatio-Temporal Map
  3. Access Ranking

In addition, access analytics maps are available at individual typhoon summary pages.

2. Method of Access Analytics

The current workflow given below is a temporary one.

  1. Remove accesses with HTTP status code other than 2XX, or with methods other than GET and POST.
  2. Remove accesses from internal network, or from robots (crawelers) included in JAIRO Crawler-List (Version 1.2).

After the cleaning, access logs are used in several ways.

Monthly Time Series
Count the number of accesses for rendering HTML content on a web browser using static or dynamic (CGI) methods. This is so-called page view. Excluded accesses consist of images and CSS, non-HTML resources such as RSS/KML/XML, JSON for particl content.
Spatio-Temporal Map
IP addresses are converted to location and visualized on the map. Distinct IP addresses are counted, which is close to visitors. Access to all resources is analyzed, including images.
Access Ranking
When the URL includes the ID of typhoons, those IDs are extracted and counted by typhoons. In addition to page views, it also includes access to non-HTML resources (KML/Atom/GeoJSON) for each typhoon.

We use MaxMind GeoLite2 for converting an IP address to a location. This method, however, has a number of limitations that cause degradation of results. First, old logs record domain names instead of IP addresses, which do no allow reverse lookup at the time of analysis. Second, some IP addresses are not included in MaxMind data. Third, old IP addresses may not be included in the latest MaxMind data.