April 20 2015

BHL Tech meeting 4/20/15
William, Mike, Trish, Bianca

BHL general
William attended Museums and the Web 2 week ago – some attendees remember the Art of Life project. Last week William and Keri conducted Workshop where he Shared BHL/SIL digitization experience with Brazilians – attendees expressed interest in sharing collections, ready to be part of BHL, if I have scanned materials how to contribute to BHL? Some attendees interested in contributing content, others expressed interest in being a member e.g. Colombia. This workshop was more technical and not policy about membership. BHL Global will be meeting in Brazil next month
Mike said Old Google code repository was ported over to github (contains current list of BHL ids, will be pushed to github everyweek) and Rewrote some exports related to move to SIL. Google code is going away.
Full text search plan –. William has not worked on this but will prioritize this now that he's back from Brazil. William will send out a detailed plan with new timeline to Bianca, Trish, and Mike

Art of Life
Science Gossip – collections that were on pause due to metadata issues are now live. Trish followup with stuff that had metadata problems and were live – who will be realigning the metadata?. Trish followup on which volumes of Gardeners Chronicle we will upload.
Jim reduced reviewers down to 10 to Zooniverse to recommend we reduce number of classifiers who need to review to each page
Mike and Trish had a call with Briana last week (heavy Zooniverse user) to discuss her image finding algorithm. We will be comparing the results of her algorithm with our algorithms and seeing which one performs the best for finding images. We will use the pages already classified in Science gossip as the gold standard. The ultimate goal is to use the best algorithm to filter the pages so Zooniverse users don’t have to look at as many pages when classifying.

Purposeful Gaming
Desmond meeting – has been working on binarization. Trish email him to see his progress.
Tiltfactor – formats finalized and files will be at page level and not books as discussed with Max in November. Meeting Tuesday at 2pm CST.
F2F meeting Set date for May12th. Confirmed there will be 10 attendees (Doug, Mike B., Trish, William, Mike L., Susan Lynch, Marty, Holly, Joe, Patrick) and reserved IT conf room
FromThePage problems – had meeting with Ben Brumfield April 9th and determined we can use FromThePage as is for now. Joe was able to find enough pages where the images are displaying correctly. Ben indicated he is working on next version of his software and several institutions are testing it. It will have support for OCR so is something MOBOT would be very interested in testing too.
Mining Biodiversity
Trish met with Riza last week to get updated on her work with the semi automated training of ARGO. Riza was able to use the previously annotated docs that Trish did and train the tool so there is not as much noise in the documents when beginning the manual annotations. Trish will focus on finishing her set of annotations in the next few weeks. Jen has begun adjudicating.
Corrected OCR group – can only meet on Saturdays
Social media – Grace pushing tools in beta and looking for feedback from TAG. Altmetric, Disqus (commenting to pages), AddThis
Trish taking off Wed and Mike will work from home Friday.