Comments on: How Embedded ChemSpider Structures Makes Blogs Searchable

By: Antony Williams

Antony Williams — Mon, 02 Mar 2009 19:15:25 +0000

Thanks to all of you for engaging in the discussion regarding the value of the embedded linking. Yes, the intention is that a new Data Source is set up for different bloggers. At least that is our direction at present. Every embedded link does result in an email for review that it is not ending up at an “inappropriate” page, kind of like the Katie Crow image that was posted to the site.

We have been watching the enormous increase in the number of links resulting from Comments, Archiving of the pages and so on. At present there is work underway to develop a way to cleanse duplicate links. Bear with us…it is a lot more difficult than you might imagine.

By: Joerg Kurt Wegner

Joerg Kurt Wegner — Mon, 02 Mar 2009 18:58:39 +0000

Agreed, I think that ChemSpider as global chemical linking repository is a great opportunity and will allow a lot of local or remote user contributions.

I checked the entries of citalopram and hexacyclinol again, and the number of my cross-links seem to increase, and the new links are rather a function calls than final links.

By: Cameron Neylon

Cameron Neylon — Mon, 02 Mar 2009 09:10:38 +0000

Jorg, I think it is precisely the point that by writing about a molecule you have become a datasource, and that there is real value in aggregating all of that material that is so exciting here. We will have a lot to discuss in the future about the quality of various datasources but I would guess to a first approximation that showing the first five in reverse page rank order would be a good start. I would guess that Tony is trying to encourage people by getting their names up in lights to start with but that the filtering of what is displayed will get gradually more ruthless in the future.

By: Jean-Claude Bradley

Jean-Claude Bradley — Sun, 01 Mar 2009 14:11:20 +0000

It is really nice. Since we’ve started to use this for the ONS solubility challenge – you get automatic links to the solubility data from ChemSpider.

By: Joerg Kurt Wegner

Joerg Kurt Wegner — Sun, 01 Mar 2009 11:07:29 +0000

Oh another observation, I was not sure how to use the feature, and changed the embedding for hexacyclinol a three times. Now I am listed three times as data source ? Though, also here, the linking is suboptimal, not pointing to the right post, but to the main blog domain.
http://www.chemspider.com/RecordView.aspx?id=16788289

By: Joerg Kurt Wegner

Joerg Kurt Wegner — Sun, 01 Mar 2009 11:03:00 +0000

I used the embedding feature in my recent blog post http://tinyurl.com/blog4nature

I made two observation. Maybe I need more education on this feature?

First, I was amazed to see that my blog is now listed under data sources. Just imagine 25 people and web-pages are doing the same, then the data source fields are exploding and I am wondering how this will be organized, grouped, or ranked in the future ? How to rank the good from the bad back-links, same for chemical vendors (which do not deliver or provide bad quality)?

Second, the link is pointing to the general blog (main domain), not to the specific blog post linking to the chemical compounds. I think this should be avoided, people need to pointed to the right information immediately !
Now, this is confusing, since the post was talking about the recent call from nature that more scientists should join the blogosphere. I join this statement and spread the word with my blog post. Beside, did I add popular scientific blog discussions, like the ones about citalopram and hexacyclinol.
The blog will look different in a month or year from now, how can users then find the ‘relevant’ information ? I do not want to cause link spam in any way, and rather would like not being cross-linked, than pointing people to confusing information.

Please let me know, if I used the right settings or if there is anything I could do for improving the linking character.