Hydra Metrics Interest Group - HydraConnect meetup

Thursday, Sept 24, 2015

University of Cincinnati - starting from scratch on metrics with a self-submission IR - interested in providing usage data to submitters. Participates in the maintenance of the NISO SUSHI standard, including the maintenance of the COUNTER schema.

Oregon State - migrating to Sufia-based repo - interested in changing reporting system. Sufia analytics uses Google Analytics - planning to build on that. Would like to report on what people want - pageviews and downloads, need to be reliable, need to pull themselves, want to geo-locate usage. Interested in reporting module - provide for example downloads of ETDs by departments

Northwestern - Avalon. Work with UX department to develop serious web-based SEO metrics project. At the beginning of internal repot, and still working to gain support from administration - would like to do that by demonstrating usage. Interested in figuring out how to analyze authenticated usages. A lot of content in ‘boutique’ individual sites - need to bring that into Fedora repo. As Avalon community manager, need more metrics to figure out how who is using the system. Download bounce-rate? What makes people stick around what helps keep them?

Temple - Hired assessment librarian who will be using metrics to gauge effectiveness - working group within library to look at SEO and exposure on social media. Working with GA and looking at Altmetrics to get smarter statistics.

NYU - exploring GA and SEO and event tracking to understand use of site features (e.g., facets, show links). Interested in referral stats - is the work we’re doing improving access through the systems we maintain? Currently collecting baselines. Are we putting our effort in the right place based on actual metrics. Corey is also interested metrics about metadata quality and completeness. Want to connect that to navigational metrics and query analysis - compare language of metadata to language of users. Also interested in collection analysis - e.g. depth, clustering.

Penn State - working with large hierarchical collections (e.g. archives) - need to understand what is in a particular deposit - e.g. admin dashboard

Trinity College Dublin - would like dashboard per user or institute.

From Hydra in a Box Focus group - view into a collection.

Stanford - Can improve metrics for published digital repositories. No way to reveal metrics to depositories, and lacking sense of how that’s holding back additional depositors. Concerned with lack of interest within the library about how the collections are used. This is a topic of broad interest to the Mellon Foundation - how to measure their success? Huge opportunity within Hydra. SEO is particularly interesting - presentation at OR2015 about Google Scholar.

A lot of technologies adjacent to us are succeeding with Google Scholar - OJS, ePrints. Maybe an opportunity for dialog with Google? Really only applies to ETDs and Articles - much (most) of our content goes elsewhere. Unclear what the future of Google Scholar is. Big Google is pushing Schema.org - Google Scholar isn’t. At Oregon State - 95% of traffic comes from Google Scholar. Will become important as more IRs to Hydra.

Schema.org and metatags. Oxford uses metatags with schema.org compliance - dramatically improved usage. Searches from UUID lead to content.

Sitemaps? UC interested in sitemap just for Google Scholar, another for traditional search.

From Hydra-in-a-Box - third party tools (Plum, Altmetrics, etc.) anyone using? ND is evaluating Plum Analytics - widgets would tie to particular people or items would would lead to related metrics. Question about what these track that we can’t just do ourselves? The way people use a lot of image content is by searching Google Images and pinning to Pinterest - easy analysis? Can harvest the data from the Pinterest API if you have an email on that domain - could just harvest and put a dashboard on top.

Is what we’re doing valuable? Fear that metrics would expose irrelevance? NW image repo comes from possession of the art image slide repo - rights issues could be limiting our exposure.

Who has an administrator who drives need for impact metrics? Bodleian and Stanford. Assessment librarian at Temple works closely with the Dean.

How to easily grab reports and visualize them? Some resources forthcoming on the wiki.

Thread on Hydra community a few months ago about visualization - will look up and link

Limited interest in doing development work in pursuit of this work - if you have development resources, look at the ElasticSearch stack (Elk, Elasticsearch logstash kabana)

Group will form working group for a survey.