Friday, February 13, 2015

Code4Lib 2015 - Day 1 Presentations

Code4Lib has a different structure than most other conferences I have been to.  Rather than multiple tracks it has a single track and that track is densely packed.  On the first day the common track had 10 20-minute presentations and a lightning round of another 10 five-minute presentations for a total of 20 presentations.  Previously I've tried doing a significant write-up for every presentation, but this is too overwhelming with a conference like this, so instead I'm going to have a single, long post that broadly covers the activity of the day, pointing out presentations that I particularly enjoyed or found useful in highlighted text.

Presentation 1: Becky Yoose (Grinnell College) - Your Code Does Not Exist in a Vacuum

This presentation began with a brief comparison of the different ways people think technology and society relate.  The first view is technological determinism, in which technological innovation drives societal change (e.g. the printing press was a technology that resulted in the Reformation).  The second view is technological constructivism, in which society drives what technologies are developed and adopted.  This viewpoint strives to explain why sometimes revolutionary technologies are not adopted or fail to bring about changes in society.


The third relationship between technology and society Ms. Yoose presented is technological somnambulism.  In this view, there is no driving force but both things seems to stumble forward like a sleepwalker, unaware of the full impact one is having on the other.


Taking up this view as the likely more correct one, Ms. Yoose went on to describe the potentially negative effects on library culture that the adoption of concepts from the world of programming have on it.  The first concept taken on in this way was that of the Fast Fail, where many ideas are tried and you quickly decide to move on when the idea fails.  Yoose pointed out that there are many costs to this model, frequently costs that institutions with limited resources may have trouble absorbing.


The second fad from the world of programming is from the world of open source where there is a hierarchy of coding circles with core contributors at the center and regular users at the outside.  This puts the skills of coders as the most valued skill with other skills and roles in a project as secondary, and end users as leeches sucking value from the work of others.  When it comes to cultures such as libraries, this is a potentially damaging way of looking at things and a view that should be questioned, modified, and/or discarded.


Yoose closed with  these questions:

  • As we make things work, what kind of world are we making?  
  • Does the world we are making match the world we want to make?
This was a rapid, dense, and challenging talk, but overall a good and thought provoking one.

Presentation 2: Sibyl Schaefer - Designing and Leading a Kick-a** Tech Team

Ms. Schaefer described the situation at institution where she was charged with take control of the "D-Team", a department of technology professionals.  She started out by asking staff outside of the D-Team about attitudes, values and priorities.  the D-Team was viewed as an insular group and there was considerable anxiety about technology and change.  She set goals for the team as providing access, custody of resources, and professionalism.

Using the calculation of "goals + values = consistency" it was important then to establish values to provide consistent service to the staff.  The team's values were set as: service oriented/user centered, a need to iterate to a great product, being in the picture for the long haul, and an ability to continuously learn.

Ms. Schaefer then provided the following recommendations for assembling and managing a team:
  • When hiring - figure out what you need.  Look for authenticity, values, curiosity, tech ability, archival background.
  • Balance making and managing.
  • In delegation - choose the right person for the task, agree on expectations, stay engaged, create accountability and learning. 
  • Get and provide feedback.
  • In communication - Individual weekly check-ins, weekly reports, monthly reports are important
Presentation 3: Erin White - Programmers Are Not Projects

Erin White followed Sibyl Schaefer with another description of managing programmers in a team.  She emphasized the need to balance hard skills, easily defined skills that require easily defined knowledge of some kind, with soft skills, those for managing people that are based on context and the people with whom you work.  She then went over a number of examples of soft skills and the way you use them to manage a team that works.

Presentation 4: Coral Sheldon-Hess - Leveling up Your Code with Code Club

Coral Sheldon-Hess described her experience with getting together with others to improve computer programming skills.  This was an interesting concept that I think might possibly have some use in some context or another in my life, but I'm not entirely sure where yet.


Code Club is a small social group, in Sheldon-Hess' example all women, who get together on a regular basis to go over some sample code.  They can find sample computer code in any number of places and a different person is responsible for running the group each week.  They spend an hour going over a bit of code and working out together what the code does and why it was assembled the way it was.


In her case she described a group that met using Google Hangouts, scheduled their meeting time using Doodle, and was in the 4-7 person range.


It sounded like a good way for people to learn how to think about writing better computer code by studying what others have done to solve a problem.


Presentation 5: Bill Levay - A Semantic Makeover for CMS Data

This was an interesting presentation made by a graduate student.  He described the project he worked on in which Javascript, SQL and Python were used to tag photographs of Jazz musicians so that they could be searched by the people in the photos, when the photos were taken and where the photos were taken.

What was most interesting to me were the new (to me) references of using the DBPedia and GeoNames URIs to link to data in different resources in a programmatic way.

Presentation 6: Jason A. Clark and Scott W. H. Young - Your Chocolate is in My Peanut Butter! Mixing up Content and Presentation Layers to Build Smarter Books in Browsers with RDFa, Schema.org, and Linked Data Topics

The presenters, from the University of Montana, described their work with digitizing books to make a result from a paper book that has more of the advantages of materials normally on the Internet, like hyperlinking to related external resources, full-text searchability, and the ability to link to specific sections within the book easily.

A couple examples of their finished product can be found here:
Home Cooking
Opsis Literary Arts Journal

Presentation 7: Anne Wooton - Helping Google (and Scholars, Researchers, Educators, & the Public) Find Archival Audio

This was an interesting session if only to learn of the site the speaker develops.  Anne Wooton is one of the founders of  PopUpArchive.com, a site that mainly works with other agencies to create text searchable audio.  Through their process, which has required a great deal of effort to refine to the point that it works now, they can get a recording in English and generate a rough transcript of the text.  The text is then associated with the locations where it occurs.  This allows you to search the audio with regular text and then you are dropped into the recording where the text occurs.

The process also creates what are essentially the audio equivalent of thumbnails, where in a list underneath the player you can see a list of times in the audio and what text is being spoken there.

To do this they are using the search tool Sphinx and the software Kaldi for speech to text.  The speech to text tool was trained using oral history and public media sources in English, because that is the kind of audio they are typically processing.  At the moment their system doesn't work with languages other than English.

Lightning Talks I

Lightning talks are 5 minute presentations given by attendees that are scheduled during the conference (rather than in advance, like all of the main presentations were).  During the conference there were three blocks of lightning talks, with each block having 9 to 10 short presentations.  The presentations in this first round of lightning talks (with only ones I was particularly interested in having any extra description beyond the title) were:
  • Automated Entity Extractions to Relate Library Resources
  • Open Source Digital Archiving Toolkit, ResCarta - This was kind of interesting if I were to have a need for the product, which I guess I might, eventually.  It is a piece of software written in Java that is a kind of one-stop-shop for processing different kinds of files.
  • Information Design Thoughts - The main thing I got out of this was a book recommendation: The Design of Everyday Things
  • Vufind & Worldcat Discovery API
  • Video Accessibility on the Web  - This was an informative and entertaining talk demonstrating very quickly how adding text tracks to a variety of video formats can be made easy by the use of a file format called WebVTT.  This is important if you want to make video accessible.  It is also an easy way to make a karaoke video, which they demonstrated by showing a segment of  the music video for Tubthumping by Chumbawamba.  This resulting in a considerably large number of people joining in on the chorus, which was pretty much where their demo video cut out.
  • Teaching Your ILS How to Accept Money for Fines - We are already doing this through our consortium and seeing a discussion of the process made me glad that someone else has taken care of it.
  • Fedora 4 Migration
  • LDPath - A demonstration of using this software to extract data about Portland from geonames.
  • Self-deposit of Scientific Data
  • Bread (How Baking Bread Made Me a Better Programmer) - Making bread made the prsenter a better programmer.  Learning to do this 1.) helped him embrace his fears 2.) learn the components (he reveres Alton Brown's I'm Just Here for More Food) 3.) Patterns are useful 4.) Patterns have to be implemented into a larger context 5.) Learn other styles

Presentation 8: Eben English - Boston Public Library - Book Reader Bingo

This presentation was different than what most people would guess it was about (I would think they would guess e-book readers, like a Kindle) or what I though it might be about (a book scanner).  In this case it was an overview of the different products that can be used within a browser to show a book, with a detailed comparison of the different features and drawbacks of each.  It was an entertaining and well-done talk which could come in very handy if I ever have a book that I need to have embedded into a site for on-screen reading (and given the fact that's kind of what we do with our newsletter nowadays, this eventuality could happen sooner than I would otherwise anticipate).

Presentation 9: Megan Kudzia and Kate Sears - Leveling Up Your Git Workflow

In this talk, the presenters described their problems with Git (and for the most part they were talking specifically about the most public implementation of Git, GitHub) and how their experience changed the way they thought about the resource.  They went on to describe how the changes in their thinking helped them use Git more effectively.

Presentation 10: Terry Brady (Georgetown University) - Got Git? Getting More Out of Your GitHub Repositories

This session had a variety of tips for using GitHub for collaborating with someone on code and for communicating or annotating code that is hosted on GitHub.  It tied in very nicely with the presentation before it.  I have used GitHub some and it is an important resource for downloading projects in development (largely supplanting for my use the older SourceForge).  There were several interesting tidbits, but I'm not sure how soon I would be using them.

No comments: