tag:blogger.com,1999:blog-17889588.post116576455827347523..comments2024-03-13T07:14:55.283+01:00Comments on chem-bla-ics: Including SMILES, CML and InChI in blogsEgon Willighagenhttp://www.blogger.com/profile/07470952136305035540noreply@blogger.comBlogger9125tag:blogger.com,1999:blog-17889588.post-23321402782605888952010-08-10T22:45:51.995+02:002010-08-10T22:45:51.995+02:00The update examples are available at:
http://chem...The update examples are available at:<br /><br /><a href="http://chem-bla-ics.blogspot.com/2010/08/xhtmlrdfa-chemical-examples.html" rel="nofollow">http://chem-bla-ics.blogspot.com/2010/08/xhtmlrdfa-chemical-examples.html</a>Egon Willighagenhttps://www.blogger.com/profile/07470952136305035540noreply@blogger.comtag:blogger.com,1999:blog-17889588.post-67751517862466261792010-08-10T21:15:39.649+02:002010-08-10T21:15:39.649+02:00Ramoonus: the NS never actually resolved to a webp...Ramoonus: the NS never actually resolved to a webpage. Namespaces are URIs, not URLs.<br /><br />About, chemicalcompound. Yeah, better use chem:compound there, which would be namespaced.<br /><br />I'm posting some more examples in my blog right now.Egon Willighagenhttps://www.blogger.com/profile/07470952136305035540noreply@blogger.comtag:blogger.com,1999:blog-17889588.post-24729760206567662632010-06-08T13:30:25.430+02:002010-06-08T13:30:25.430+02:00span class="chemicalcompound">asperin...span class="chemicalcompound">asperin<br />Isn`t such a strange idea<br />but putting it into line with the others gives something like:<br /> span class="chem:compound">asperin.Anonymoushttps://www.blogger.com/profile/16784623066903654688noreply@blogger.comtag:blogger.com,1999:blog-17889588.post-19985106612621157842010-06-07T09:46:58.456+02:002010-06-07T09:46:58.456+02:00The XMLNS isn`t working anymore (error 404)The XMLNS isn`t working anymore (error 404)Anonymoushttps://www.blogger.com/profile/16784623066903654688noreply@blogger.comtag:blogger.com,1999:blog-17889588.post-21631102606718624152007-02-19T14:43:00.000+01:002007-02-19T14:43:00.000+01:00Richard,some people put the InChI in the @alt attr...Richard,<BR/><BR/>some people put the InChI in the @alt attribute when giving an image, but this lacks semantics.<BR/><BR/>To get it indexed by Google, you might want to try to put it in a keyword in the header:<BR/><BR/><META NAME="keywords" CONTENT="InChI=1/bla, InChI=1/bla2, etc"><BR/><BR/>Never tried that myself yet, though...<BR/><BR/>PS. you blogger profile is not public, so can't see to which blog the name 'Richard' belongs...Egon Willighagenhttps://www.blogger.com/profile/07470952136305035540noreply@blogger.comtag:blogger.com,1999:blog-17889588.post-16338809718653847712007-02-18T18:17:00.000+01:002007-02-18T18:17:00.000+01:00Hi Egon,Very nice work. What if I don't want the I...Hi Egon,<BR/><BR/>Very nice work. What if I don't want the InChI to actually be visible on the page, but just indexed by Google, CB, etc.?Rich Apodacahttps://www.blogger.com/profile/09995006481691926615noreply@blogger.comtag:blogger.com,1999:blog-17889588.post-1167814962224273262007-01-03T10:02:00.000+01:002007-01-03T10:02:00.000+01:00Hi Mark,Thanx for your comments. I am now actively...Hi Mark,<BR/><BR/>Thanx for your comments. I am now actively promoting RDFa to do this semantic chemistry and have extended an aggregator to detect those semantics:<BR/><BR/>http://wiki.cubic.uni-koeln.de/cb/inchis.phpEgon Willighagenhttps://www.blogger.com/profile/07470952136305035540noreply@blogger.comtag:blogger.com,1999:blog-17889588.post-1165788455803883852006-12-10T23:07:00.000+01:002006-12-10T23:07:00.000+01:00Hi Egon,This is all very interesting. :)I know not...Hi Egon,<BR/><BR/>This is all very interesting. :)<BR/><BR/>I know nothing about chemistry, but I think I can see enough about what you are trying to do to say that RDFa is way out in front for your needs...contrary to your view that:<BR/><BR/>"[T]his [using RDFa] is more advanced, and currently does not have practical advantages over the use of microformats."<BR/><BR/>The first advantage is that terms are namespace qualified. There are only two situations I can think of where this would be of <B>no</B> use to you, and both of them seem very unlikely to me.<BR/><BR/>The first is that your blogs, and your colleagues' blogs, only ever talk about chemistry, i.e., you won't ever want to use terms from other disciplines. The second, is that no-one else will ever use your terms.<BR/><BR/>The former seems very likely to me, since your terms would surely be useful in blogs about medicine and physics at the very least? On the latter, do you really think that no other discipline or organisation will use the unqualified term 'smiles'? :) If they do, then your search results will go back to being as bad as they probably are now.<BR/><BR/>(Like I said, I know nothing of chemistry, but I'm assuming that a Google search for "acetic acid" is pointless for you guys...is that right?)<BR/><BR/>So if you were going to, how should you use namespaces? Your suggestion when illustrating RDFa is that you have one prefix, and then use @class to set 'types'<BR/><BR/>"* SMILES: <span class="chem:smiles">CCO</span><BR/> * for CAS registry numbers: <span class="chem:casnumber">50-00-0</span><BR/> * for InChI: <span class="chem:inchi">InChI=1/CH4/h1H4</span>"<BR/><BR/>However, in the world of RDF, you generally want each organisation to 'own' its own terms, or its own taxonomies. So I would have thought you'd want to have terms like this:<BR/><BR/> smiles:cco<BR/> cas:50-00-0<BR/> cid:176<BR/><BR/>SMILES, CAS and CID would have namespace URLs of their own, and each organisation would define its terms.<BR/><BR/>Using the last one as an example ('cid:176'), in RDFa you would be able to use such a term like this:<BR/><BR/>I had run out of <span content="cid:176">acetic acid</span>, but luckily Egon had some <span content="cid:176">Natriumacetat</span>.<BR/><BR/>(I told you I know nothing of chemistry...I just perused the NCBI site to make up some examples! But I hope the point being made is clear.)<BR/><BR/>Note that I've used the attribute @content which comes form <meta>, but is made more widely available by RDFa. This allows you to give a precise term for whatever is in the element, and the classic example we usually give is something like:<BR/><BR/>Today the <span content="people:TonyBlair">Prime Minister</span> flew to the <span content="country:usa">US</span> for talks.<BR/><BR/>But whilst RDFa can be very simple (as these exampled show), it also opens up the possibility of providing far more metadata. For example, the NCBI site could embed RDFa in its pages, and then an RDFa processor could pluck out the information and make use of it:<BR/><BR/><div about="cid:176">Acetic acid has the same parent compound as <a rel="ncbi:parent" href="[cid:11954357]">Cupric acetate</a>.</div><BR/><BR/>(I have no idea whether that's completely non-sensical. :) I just tried to find some relationships between items to illustrate the point.)<BR/><BR/>Regards,<BR/><BR/>MarkAnonymousnoreply@blogger.comtag:blogger.com,1999:blog-17889588.post-1165777806919511462006-12-10T20:10:00.000+01:002006-12-10T20:10:00.000+01:00Hi,Tagged SMILES added as you suggested, many than...Hi,<BR/><BR/>Tagged SMILES added as you suggested, many thanks<BR/><BR/>http://homepage.mac.com/swain/Sites/CMC/News/files/ab724be749104d0ef80af00d3db8cc63-8.htmlAnonymousnoreply@blogger.com