Emboldened by the experimental spirit of the Books in Browsers community, JEP’s BiB IV Proceedings issue charts new territory for our editorial and production processes, challenging what we traditionally thought of as conference proceedings. By reducing a proceedings to recorded video, slides, and making submission of papers optional, this lighter proceedings narrows the time gap between conference closing and proceedings publication. However, in doing so introduces a new set of editorial and production challenges to our traditional workflows. Where will videos live? How will we (should we?) produce transcripts of video? What to do if a presenter elects not to submit a paper? What level of copyediting should we apply to submitted papers? What about additional conversations and commentary that happens on the web? In a conference-y spirit of sharing, here’s a closer look at how we addressed these questions and went about putting together the BiB IV Proceedings.

The Model

In designing a proceedings model for BiB IV, we made paper submission optional and devised a fallback method if a paper was not submitted. For example, if a BiB conference author/presenter chose not to produce an article-length version of his or her talk, we are embedding the recorded oral presentation, creating an edited transcription, and including this along with presentation slides (if provided). If transcription is not possible or a recording is not available, at the least an abstract and slides are provided. In this way, and with only a light copy editing review on submitted papers, we hoped a proceedings issue could be assembled fairly quickly (4 months after conference conclusion) while still expanding and accurately representing the ideas and dialog of the conference with the wider world.

Since many presenters weren’t submitting a paper, transcripts of talks became a key component of the proceedings – something we believe will be important to this issue. Transcripts not only help serve as an indexable and searchable record of the BiB IV talk for search engines and our internal site search, but are good for long-term preservation of the BiB IV conference. Perhaps most importantly, transcription provides an accessible user experience for readers with hearing disabilities or for those who would rather read than watch or listen to a presentation.

The Process

With an editorial model devised, a real challenge was figuring out how to execute the proceedings. Incorporating video into the proceedings posed one major challenge for JEP’s production workflow – how should we affordably generate sensible transcripts?

Because BiB IV was already set to be live streamed, most of the streamed recordings were later uploaded to BiB friend Publishing Perspectives’s YouTube Channel and organized into a YouTube playlist. Access to this channel was given to us through Google+’s multiple admins functionality. We used this access to collect videos and their transcripts and upload revised transcripts.

The transcribing and quality control were labor-intensive processes, but as mentioned earlier, a worthwhile investment for both JEP readers and the long-term preservation of the issue. To help speed up the transcription process, we relied on YouTube’s auto-generated transcripts to create a rough outline, which we would later edit. While helpful, YouTube’s auto-transcription technology has a long way to go. For example, oddly, one talk given in English from a presenter with a French accent was interpreted and auto-transcribed in French only. In other cases, words and phrases were completely or partially misidentified: “importance of libraries” as “important survivor”, “publishing” as “polishing”,  “print on demand” as “print demands”, or “let us tailor” as “latest hello”. In addition, to ensure a pleasurable and usable reading experience, pauses in speech such as “um...” and “ah...” were removed from transcripts. In this regard, auto-generated transcripts required a close reading and editing process prior to publication (Many, many thanks to Michigan Publishing’s Digital Publishing Coordinator Kelly Witchen for her work on this). For videos longer than twenty minutes, we learned a transcript was not auto-generated by YouTube and therefore transcripts had to be manually generated (and many thanks to University of Michigan School of Information graduate student Kathryn Horne for her hours spent creating these).

Transcripts aside, we also thought it important to capture conversations during and immediately after the conference. We’ve organized this activity into commentaries, summaries, responses to specific talks, supplementary materials, and photos in an index of links at the end of the issue.

Looking Ahead

With a model and a process for editing and production of an experimental proceedings issue under our belts, we now look forward to seeing how our readers interact with the issue, finding a new set of questions to ask (and potentially report back on): Do readers give preference to content types (papers, video, slides)? Will embedded video recordings be watched? For their entire duration? Will full transcriptions be read? Slide decks downloaded? Will the timeliness of the content increase the chance that readers will comment on articles? What will usage and interaction with the proceedings look like over time? Will they be cited? Can a published proceedings help BiB reach a wider audience and increase its already thriving community? Answers to these questions and more will help us improve the experiment to publish relevant, timely and innovative conference proceedings for our readers in the future.