Feb 2 2015

Tech meeting 2/2/15
Mike, Trish, William, Bianca
BHL general
Question from Bianca – is there a way to view darkened content in IA? Is that part of metadata? Yes in meta.xml file, notes part of approval element. Would like to periodically run a report.
Bianca talking with Alvin about DO that need to be removed for materials that were unpublished for various reasons. Asked him process of getting rid of DOIs. Deleted DOIs have message from crossref.
Institutional collections discussion with collections committee – uncomfortable letting go of it because you can’t search on institution in advanced search. Want to search within a particular contributor - is that easy to do? Mike says no its not easy – it would require both change to UI and index since we don’t currently index the contributor information
Contributor Browse – got some suggestions on how to improve this page. Can Bianca submit as Gemini ticket with low priority?
Mike wonders what about Harvesting publisher data from 264 field? Bianca can submit as Gemini issue with low priority

Art of Life
Promoting extra 350k BHL images in Flickr – blog post, twitter and FB posts. Blog post recommended format taxonomy:common=value
Trish wondered if someone mistypes the word “taxonomy” or “binomial”? Does it get missed by EOL? Bianca thinks probably but not sure. Maybe not a big deal since Mike reports that of the 25k tags with that format only 15 or so have been misspelled. If we want we can fix those tags ourselves (Trish send list to Bianca)
Grace says the post is Getting lots of attention- tweet got picked up quite a bit this morning- now over 400 interactions and seen over 23,000 times
Grace would like to know if she could get monthly updates of how much people are tagging and she will incorporate that info into her tweets? Mike will look into how easy it is to query the API for particular machine tags. Trish could check with British Library folks about if they know of tools for doing this easily.
Trish Asking MBG to do a press release.
Tagging parties – worthwhile? Could William do an image tagging party at MW2015 in March?

ConSciCom/Zooniverse –
Moving along. Looked at current version of beta and made some suggested tweaks to UI. The 5 journals Mike sent should go live this week. He will begin to upload the other 15 journals.
Live date set for mid-March (exact date TBD). Beta testing in next 2-3 weeks.
Getting content back out from Zooniverse – would be good to test out getting their exports.

Purposeful Gaming
mediaWiki is to be done and we need to test out moving files over. Mongodb setup and working. Once TEsseract working Mike will be pushing files there.
In comparing outputs Mike will be using 15 for accuracy score and .7 Darwin Score.
Things that fail those thresholds will get kicked out for transcription or manual review.
Next steps with generating better OCR – Mike L will start generating Tesseract outputs. He will use the criteria of English publications from 1900-1923. Mike did a query - 8k items. Mike will test out 100 items . A lot of these will be seed catalogs
Tiltfactor – 2nd round of testing. Trish identified 7 people to test. Mix of BHL users interested in OCR correction sent from Jackie, Grace and Bianca and PG staff
Text to image linking tool – what do we want to do about desmond’s email? He said he is hoping to develop more by June 2015 but could do sooner and more to our specifications he we could pay him for the work? Trish Asked him a ballpark figure for the work. Mike wondered if we’ve missed our window of opportunity? What’s the latest we could get it from Desmond? End of March? What’s the latest we could get it to Tiltfactor? Lets wait to hear from Desmond on timeline and money to see if that would be feasible.

Mining Biodiversity
Annotations – Trish and William annotating golden standard now.
Social media – Grace and Anatoly look at use of our big picture. Look at followers of our followers. Will need programmer dedication to implement functionality for admin dashboard and visual keys for users "to promote discussions on BHL objects".