http://www.sciencebase.com/science-blog/pubchem-statistics.html Science Blog from Freelance Science Writer David Bradley Sat, 20 Mar 2010 17:40:20 +0000 http://wordpress.org/?v=2.9.2 hourly 1 http://www.sciencebase.com/science-blog/pubchem-statistics.html/comment-page-1#comment-380965 Egon Willighagen Fri, 06 Jul 2007 07:16:46 +0000 http://www.chemspy.com/chemistry-news/1459.html#comment-380965 For log analysis, I can very highly recommend Lire (see logreport.org, disclaimer: I am a former developer). It allows to define custom analysis and is very flexible in general. The software basically works with an intermediate DLF format, where each column contains a particular bit of information of interest. For example, a log file processor can easily add a column with the type of request, extracted from the URL. Such a processor could also look at the request and extract the CID, possibly by reverse lookup of InChI and SID. A report generator can then do all sorts of analysis and, for example, create a report for 'Most looked up Compounds per Continent', or 'Most Common Set of Five Compounds' by tracking the click-path people used. All quite trivial in Lire. The good news, Lire is distributed with several Linux distributions. The bad news, development has slowed down somewhat. For log analysis, I can very highly recommend Lire (see logreport.org, disclaimer: I am a former developer). It allows to define custom analysis and is very flexible in general. The software basically works with an intermediate DLF format, where each column contains a particular bit of information of interest. For example, a log file processor can easily add a column with the type of request, extracted from the URL. Such a processor could also look at the request and extract the CID, possibly by reverse lookup of InChI and SID. A report generator can then do all sorts of analysis and, for example, create a report for ‘Most looked up Compounds per Continent’, or ‘Most Common Set of Five Compounds’ by tracking the click-path people used. All quite trivial in Lire. The good news, Lire is distributed with several Linux distributions. The bad news, development has slowed down somewhat.

]]> http://www.sciencebase.com/science-blog/pubchem-statistics.html/comment-page-1#comment-380964 Antony Williams Fri, 06 Jul 2007 01:56:56 +0000 http://www.chemspy.com/chemistry-news/1459.html#comment-380964 I'm interested to know where there are any available stats on the number of users per day querying PubChem, the number of searches they execute and the types of searches they do. In fact, it would be interesting to know similar stats for CAS Scifinder searches. I have to imagine that the number of queries must be in the tens of thousands per day for Scifinder and thousands per day for PubChem...but these are just perceptions...are there actual numbers anywhere? I’m interested to know where there are any available stats on the number of users per day querying PubChem, the number of searches they execute and the types of searches they do. In fact, it would be interesting to know similar stats for CAS Scifinder searches. I have to imagine that the number of queries must be in the tens of thousands per day for Scifinder and thousands per day for PubChem…but these are just perceptions…are there actual numbers anywhere?

]]>