Lemon does not read whole document

[jaxim]

Hello,
we have a problem with with Lemon 8. If we upload a document (till now we are just trying the sample) Lemon does not read it for 100%. There is a variation in percents (5-96 but never 100%) if we upload it on different computers (different internet connection). The 4% which Lemon8 does not read are citations.
Thanks for any help in advance
Jakub

[jerico]

Hi Jakub,

can you maybe post a screenshot so that I better understand what you mean by "not reading 100%"? Is this maybe a connection or a timeout problem? Or do you mean that the document is not fully transformed into XML?

Thanks,
Florian

[jaxim]

Hi florian,
thanks for answer, here are screenshots of our problem...I hope it helps...
Jakub

[jerico]

Hi Jakub,

thanks for the screenshots. This makes it much easier for me to see what your difficulty is.

What you're experiencing there are not upload errors. The percentages simply show how many of the citations and/or document parts have been successfully split up into fields, paragraphs, etc (or for a single citation: how well this citation could be interpreted by Lemon8) as required by the NLM standard. It shows you how "confident" Lemon8 is in the results that have been generated so far.

The variations you experience have most probably to do with some temporary downtime of the external services that Lemon8 relies upon. We recently had trouble with the ParsCit and FreeCite web services which went on and off all the time. Unfortunately PKP (and thus Lemon8) has no control over these external services. You have to know that Lemon8 is something like a "meta citation parser" that uses third party web services in the background. If these are temporarily not available then your results will deteriorate.

Even with all web services up and running it is very improbable that you get a result of 100% for any document. 100% would mean that all citations have been found in an external database (and the document could be parsed into XML fully autonomously) and that's virtually impossible. This means that you'll always have to go into the editor and fix the citations and/or document parts that were not correctly split up by hand.

Once you've done that you can advance to the export tab and generate NLM-compliant XML.

BTW: Lemon8 is no longer being actively developed by PKP. It is an experimental tool. We're currently working on integrating all Lemon8 functionality into OJS. The upcoming OJS release will most probably contain a beta-version of the Lemon8 citation parser with a revamped user interface and with lots of extension points for your own citation services and database connectors.

For full XML output support in OJS you'll have to wait a bit more. We're planning to release that in 2011.

Hope this helps a bit to clear things up,

Florian

[jaxim]

Hi Florian,
thanks for help and Your time You spent on it.
Looking forward for new version .
Jakub