Some statistics about nmrshiftdb2

I compile statistics about usage of the web interface from time to time and I thought it might be nice to share these. So here are a few statistics and comments. Just to keep users assured that privacy is respected: No queries, personal data or IPs are published and no information which can be put together to build profiles are made public. This is just aggregate statistics we use here. Logging of queries is in line with the nmrshiftdb2 privacy policy outlined at http://nmrshiftdb.nmr.uni-koeln.de/portal/js_pane/P-Help?URL=using.html#register

Firstly I compliled statistics of the overall usage, taken from the queries actually executed. So these are not http access statistics, which tend to be cluttered by robots etc., but actual queries executed. Included are predictions and all types of searches. Some searches have their own counting, the rest is in other. The first image shows the usage ove time. These include all NMRShiftDB and nmrshiftdb2 servers, so when there were several nodes all statistics were put together. For the absolute figures, I put in the last few months in the following table:

Year|Month|Prediction|Exact Substructure|Spectrum|Fuzzy Substructure|Name|CAS|Other
----|-----|----------|-----|------------|--------|------------------|----|---|-----
2014| 7| 1431| 283| 420| 149| 114| 114| 1739
2014| 8| 1271| 247| 375| 80| 76| 64| 1860
2014| 9| 1232| 408| 407| 138| 142| 169| 2333
2014| 10| 1471| 516| 402| 123| 153| 96| 3493
2014| 11| 1088| 680| 443| 142| 142| 120| 4499
2014| 12| 913| 366| 314| 149| 101| 82| 3148
2015| 1| 1360| 331| 281| 99| 64| 88| 3172

In order to find out more about where our users came from, I used whois command line tool for the countries of the orginasations, who were assigned the IPs, which were used for database queries in December 2014 or January 2015. The result is shown in the figure below.

A full list of all countries and the number of their requests is below:

US|NL|DE|AU|RU|IN|CA|CN|GB|IT|FR|ES|PL|SE|FI|TR|BY|TZ|
--|---------|--|-----------|--|----|--|----|--|---|--|----|--|----|--|----|--|----|--|----|--|--|--|-----|--|--|--
658|519|410|189|140|121|105|81|75|65|64|29|28|26|25|24|23|22|

KR|UA|BR|th|IR|AT|JP|MU|TW|ID|EG|JO|IE|MY|BE|MX|
--|---------|--|-----------|--|----|--|----|--|---|--|----|--|----|--|----|--
21|19|19|17|17|16|16|15|15|14|12|12|11|11|10|8|

BG|ZA|TH|SI|CH|HU|cn|AP|OM|KZ|CO|LK|tr|HR|VE|UY|
--|---------|--|-----------|--|----|--|----|--|---|--|----|--|----|--|----|--
8|7|7|7|7|5|5|4|4|4|4|4|3|3|3|3|

DK|NZ|GR|VN|EU|IQ|MO|GF|SY|HK|RO|AR|IL|IS|BA|
--|---------|--|-----------|--|----|--|----|--|---|--|----|--|----|--|----|--
3|3|3|2|2|2|2|2|2|1|1|1|1|1|1|

PT|vn|CZ|CY|SG|CU|SK|CR
--|---------|--|-----------|--|----|--|----|--|---|--|----|--|----|--|----|--
1|1|1|1|1|1|1|1

The country codes are resolved at https://www.ripe.net/membership/indices/. Also note the reliability of the data is not 100% since the definition of "country" in the contact information of the domain registrars it not clear and may be outdated. Also some addresses do not reveal actual information, but point to other registrars (RIPE being based in the Netherlands is partly responsible for the many NL entires). We still can see some patterns: US requests lead, then come the Netherlands and Germany. Next are Australia, Russia, India, Canada and China. A block of Europena countries (UK, Italy, France, Spain, Poland, Sweden, Finland, Turkey and Belarus) occupy the next places. From there (with low user numbers in comparision) we have countries from all over the world. Judging from this, there is no clear geographic pattern. The populous and economically strong countries make most of the users, but virtually all of the world shows up.

I also had a look at the hostnames, as far as they could be retrieved by using nslookup. The most frequent organisations are actually internet service providers. 32 hosts belong to German universities (defined by having ".uni-" in their hostnames), 58 are ending with .edu, so are presumably American universities. 49 hosts come from .ac.*, which should be universities in several countries including e. g. the UK or Austria. If we look at top level domains, we get these numbers of hosts ending with the respective TLD: .nl 6, .de 120, .au 3 (a lot of "Australian" requests point to APNIC, based in Australia), .ru 10, .in 12, .ca 22, .cn 0 (it seems Chinese providers do not produce nslookup results), .uk 17, .it 20 and .fr 20. .com is 45 and .net 126. It looks like there is little connections between the country information in whois and the TLD. A lot of these are due to providers, which often use .net TLDs, and some to incomplete information in whois.