The Concept

I have had some ideas I wanted to try out for using the iPad as a tool for collecting photo metadata. Working in a corporate archive, I have become aware of several collections of photos without much metadata associated with them.

The photos are the property of (or are in the custodial care of) the company I work at (in their corporate archive).

The subject of the photos are either of two general areas:

1. The minority language speaking people groups that the employees of a particular company worked with, including anthropological topics like ways of life, etc.
2. Photos of operational events significant to telling the story of the company holding the photos.

Archives in more modern contexts are trying to show their relevance to not only academics, but also to general members of communities. In the United States there is a whole movement of social history. There are community preservation societies which take on the task of collecting old photographs and their stories and preserving, and presenting them for future generations.

The challenge at hand is: "How do we enrich photos by adding metadata to photos in the collections of archives?" There are many solutions to this kind of task. The refining, distilling, and reduction of stories and memories to writing and even to metadata fields is no easy task, nor is it a task that one person can do on their own. One solution, which is often employed by community historians is the personal interview. By interviewing the photographers or people who were at an event and asking them questions about a series of photos it presents an atmosphere of inquisitiveness and one where the story-teller is valued because they have a story-listener. This basic personal connection allows for interactions to occur across generational and technological barriers.

The crucial question is: "How do we facilitate an interaction which is positive for all the parties involved?" The effort and thinking behind answering this question has more to do with shaping human interactions than with anything else. We are also talking about using technology in this interaction. This is true UX or (User Experience).

Past Experience

This past summer I have had several experiences with facilitating one-on-one interactions between knowledgeable parties working with photographs and with someone acting on behalf of the corporate archive. To facilitate this interaction a GoogleDoc Spreadsheet was set up and the person acting on the behalf of the archive was granted access to the spreadsheet. The individual conducting the interview and listening to the stories brought their own netbook (small laptop) from which to enter any collected data. They were also given a photo album full of photos, which the interviewee would look through. This set-up required overcoming several local environmental challenges. As discussed below, some of these challenges were better addressed than others.

Association of Data to a Given Photo

The challenge of keeping up to 150 photos organized durring an interview so that metadata about any given photo could be collected and associated with only that photo. This was addressed by adhering an inventory sticker to the back of each photo and assigning each photo a single row in the GoogleDoc Spreadsheet. Using GoogleDocs was not the ideal solution, but rather than a solution of some compromises:

Strengths of GoogleDocs

• One of the great things about GoogleDocs is that the capability exists for multiple people to edit the spreadsheet simultaneously.
• Another strength of GoogleDocs is that there is a side bar chat feature so that if there is a question durring the interview that help could be had very quickly from management (me, who was offsite).
• The Data can be exported in the following formats: .xlsx , .xls , .csv , .pdf.
• There was no cost to deploy the technology.
• It is accessible through a web-browser in an OS neutral manner.
• The document is available wherever the internet is available.
• A single solution could be deployed and used by people digitizing photos, recording written metadata on the photos, and gathering metadata during an interview.
• Most people acting on behalf of the archive were familiar with the technology.

Pitfalls of GoogleDocs

• More columns exist in the spread sheet than can be practically managed (The columns are presented below in a table). There are about 48 values in a record and there are about 40,000 records.


More columns than can be practically managed

• Does not display the various levels of data as levels of data as levels in the user interface.
• Cannot remove unnecessary fields from the UI of various people. (No role-based support.)
• Only available when there is internet.

Maximizing of Interview Time

To maximize time spent with the interviewee the photos and any metadata written or known about a photo was put into the GoogleDoc Spreadsheet prior to the interview. Sometimes this was not done by the interviewer but rather by someone else working on behalf of the archive. Durring the interview the interviewer could tell which data fields were empty by looking for the gray cells in the spreadsheet. However, just because the cells were did not mean that the interviewee was more prone to provide the desired, unknown, information.

Data Input Challenges

One unanticipated challenge which was encountered in the interviews was that as the interviewer would bring out an album or two of photos that the interviewees would be able to cover more photos than the interviewer could record.

Let me spell it out. There is one interviewer and two interviewees there are 150 photos in an album lying open on the table. All three participants are looking at the photo album. The interviewee A says look that is so-and-so and then interviewee B (because the other page is closer to them) says and this is so-and-so! This happens for about 8 of the 12 facing photos. Because the interviewer is still typing the first name mentioned they ask and when do you think that was? But the metadata still comes in faster, as the second interviewee did not hear the question and the first one did but still thinking. The bottom line is that more photos are viewed and commented on faster than can be recorded.

Something that could help this process would be to in some way to slow-down (or moderate) the ability of the interviewee(s) to access the photos. Something that could synchronize the processing times with the viewing times. By scanning the photos and then displaying them on a tablet it slows down the viewing process and integrates the recording of data with the viewing of photos.

Positional Interaction Challenges

An interview is, at some level, an interaction. One question which comes up is How does the technology used affect that interaction? What we found was that a laptop usually was situated between the interviewer and the interviewees. This positioned the parties in an apposing manner. Rather than the content becoming the central focus of both parties, the content was either in front of the interviewer or in front of the interviewees. A tablet changes this dynamic in the interaction. It brings both parties together over a single set of content, both positionally and cognitively. When the photo is displayed on the laptop, the laptop has to be rotated so that the interviewees can see the image and then turned so that the interviewer can input the data. This is not the case for a tablet.

Content Management Challenges

When Paper is used for collecting metadata it is ideal to have one piece of paper for each photo. Sometimes this method is preferable to using a single computer. I used this method when I had a photo display and about 20 albums and about 200 people all filling out details at once.

People came and went as they pleased. When someone recognized someone or someplace they knew, they wrote down the picture ID and the info they were contributing along with their name. However, carrying around photo albums and paper there is the challenge of keeping all the photos from getting damaged, and maintaining the order of the photos and associated papers.

Connectivity Challenges

When there is no internet there is no access to GoogleDocs. We encountered this when we went to someone's apartment, expecting interent because the interent is available on campus and this apartment was also on campus. Fortunately we did have a back up plan and paper pen was used. But this means that we now had to type out the data, which was written down on the paper; in effect doing the same recording work twice.

Size of Devices

Photo albums have a certain bulk and cumbersome-ness which is multiplied when carrying more than one album at a time. Add to this a computer laptop and one might as well add to the list of required items, a hand truck with which to carry everything. A tablet is all in all a lot smaller and lighter.

Proof of Concept Technology

As I mentioned before, I had an iPad in my possession for a few days. So to capitalize on the opportunity, I bought a few apps from the app store, as I mentioned that I would and tried them out.

Software which does not work for our purposes

Photoforge2
The first app I tried was Photoforge2. It is a highly rated app in the app store. I found that it delivered as promised. One could add or edit the IPTC and EXIF metadata. One could even edit where the photo was taken with a pin drop interface.

Meta Editor
Meta Editor, another iPad app, which was also highly acclaimed performed task almost as well. Photoforge2 had some photo editing features which were not needed in our project. Whereas Meta Editor was focused only on metadata elements.

After using both applications it became apparent that neither would work for this project for at least two reasons:

1. Both applications edit the Standards based IPTC and EXIF metadata fields in photos. We have some custom metadata which does not fit into either of these fileds.One aspect of the technology being discussed, which might be helpful for readers to understand, is that these iPad applications actually embed the metadata into the photos. So when the photos are then taken off of the iPad the metadata travels with them. This is a desirable feature for presentation photos.
2. Even if we do embed the metadata with these apps the version of the photo being enriched is not the Archival version of the photo it is the Presentation version of the photo. We still need the data to become associated with the archival version of the photo.

Software with some really functional features

So we needed something with a mechanism for capturing our customized data. Two options were found which seemed to avail themselves as suitable for the task. One was ideal the other rapidly deployable. Understanding the iPads' place in the larger place of corporate architecture, relationship to the digital repository, the process of data flow from the point of collection to dissemination, will help us to visualize the particular challenges that the iPad presents solutions for. Once we see where the iPad sits in relationship to the rest of the digital landscape I think it will be fairly obvious why one solution is ideal and the other rapidly deployable.

Placement in the iPad in the Information Architecture Picture

In my previous post on Social Metadata Collection I used the below image to show where the iPad was used in the metadata collection process.

Since that time, as I have shown this image when I talk about this idea, I have become aware that the image is not detailed enough. Because it is not detailed enough it can lead to some wrong assumptions on how the iPad use being proposed actually works. So, I am presenting a new image with a greater level of detail to show how the iPad interacts with other corporate systems and workflows.

There are several things to note here:

• Member Disporia as represented here is not just members, it is their families, the people with whom these members worked, it is the members currently working and it the members living close at hand on campus, not just in disporia.
• It is a copy of the presentation file which is pushed out to the iPad or the website for the Member Disporia. This copy of the file does not necessarily need to be brought back to the archive as long as the metadata is synced back appropriately.
• The Institutional Repository for other corporate items is currently in a DSpace instance. However, it has not been decided for sure that photos will be housed in this same instance, or even in DSpace.

That said, it is important that the metadata be embedded in the presentation file of the image, as well as accessible to the Main container for the archival of the photos. The metadata also needs to sync between the iPad application and the Member Diaspora website. Metadata truly needs to flow through the entire system.

FileMaker Pro with File Maker Go

FileMaker Pro is a powerful database app. It could drive the Member Disporia website and then also sync with the iPad. This would be a one-stop solution and therefore and ideal solution. It is also complex and takes more skill to set up than I currently have, or I can currently spare to acquire. Both FileMaker Pro and its younger cousin Bento enable Photos to be embedded in the actual database.Several tips from the Bento forums on syncing photos which are part of the database:
Syncing pictures from Bento-Mac to Bento-iPad
Sync multiple photos or files from desktop to IPad
This is something which is important with regards to syncing with the iPad. To the best of my knowledge (and googling) no other database apps for the iPad or Android platforms allow for the syncing of photos within the app.

Bento
Bento is the rapidly deployable option.What are the differences between Bento 4 for Mac, Bento for iPad 1.1.x, and Bento for iPhone/iPod touch 1.1.x?
It took me about 2 hours (while doing other stuff) to download a trial version, find out how it worked, import my data from the GoogleDoc and then sync my database with the iPad.

Here is a YouTube video demonstrating my proof of concept using Bento.

httpv://youtu.be/_Eo5Ru0BF-k

Here is a series of iPad Screen shots.

Some outstanding issues

• Geo-location of Photos in Bento. Bento version 4 does have location fileds which can be used with a pin drop interface to add location data to the appropiate fileds in the database. My proof of concept demo does not demonstrate this feature.Using Geo-location fields in Bento: Working with Location Fields in Bento
  How to use Location fields in Bento for iPhone/iPad 1.1.1
• Rapid reuse of data. Because the interview process naturally lends itself to eliciting the same kind of data over a multitude of photos a UX/UI element which allows the rapid reuse of data would be very practical. The kinds of data which would lend themselves to rapid reuse would be peoples' names, locations, dates, photographer, etc. This may mean being able to query a table of already input'd data values with an auto-suggest type function.

Custom iPad App

Of course there is also the option to develop a custom iPad app for just our purposes. This entails some other kinds of planning, including but not limited to:

• Custom App development
• Support plan
• Deploy or develop possible Web-backend - if needed.

Kinds of custom metadata being collected.

The table in this section shows the kinds of questions we are asking in our interviews. It is not only provided for reference as a discussion of the Information Architecture for the storage and elements of the metadata schema is out of the scope of this discussion. The list of questions and values presented in the table was derived as a minimal set of questions based on issues of Image Workflow Processing, Intelectual Property and Permissions, Academic Merit and input from the controlled vocabulary's Caption and Keywording Guidelines which is part of their series on metalogging. The table also shows corresponding IPTC, and EXIF data fields. (Though they are currently empty because I have not filed them in.) Understanding the relationships of XMP, IPTC, and EXIF also help us to understand why and how the iPad tool needs to interact with other Archiving solutions. However, it is not within the scope of this post to discuss these differences.Some useful resources on these issues are noted here:

• Photolinker Metadata Tags has a nice display outlining where XMP, IPTC and EXIF data overlap. This is not authoritative, but rather practical.
• List of IPTC fields: List of IPTC fields. However, a list is not enough we also need to know what they mean so that we know that we are using them correctly.
• EXIF and IPTC Header Comments. Here is another list of IPTC fileds. This list also includes a list of list of EXIF fileds. (Again without definitions.)
• Various programs and applications also add their own metadata fields in the IPTC section. Here is a mapping of some of the most popular ones: http://www.controlledvocabulary.com/imagedatabases/iptc_core_mapped.pdf
• IPTC Standard Photo Metadata http://www.iptc.org/std/photometadata/documentation/IPTC-PLUS-Metadata-Panel-UserGuide_6.pdf
• Doublin Core with Photographs: http://makeit.digitalnz.org/askaquestion/questions/26
• Dublin Core Metadata Element Set, Version 1.1: http://dublincore.org/documents/dces/
• DCMI Type Vocabulary: http://dublincore.org/documents/dcmi-type-vocabulary/
• Describing Digital Content: http://makeit.digitalnz.org/guidelines/describing-digital-content/

It is sufficient to note that there is some, and only some overlap.

Metadata Element	Purpose	Explanation	Doublin Core	IPTC Tags	EXIF Tags
Photo Collection	This is the name of the collection in which the photos reside
Sub Collection	This is the name of the sub collection in which the photos reside
Letter of Collection	Each collection is given an alpha character or a series of alpha characters, if the collection pertains to one people group then the alpha characters given to that collection are the three digit ISO 639-3 code
Who input the Meta-data	This is the name of the person inputting the metadata
Photo Number	This is the number of the photo as we have inventoried the photo
Negative Number	This is the number of the photo as it appears on the negative (film strip)
Roll	This is the ID of the Roll	Most sets of negatives are cut into strips of 5 or less this allows us to group these sets together to ID a “set” of photos
Section Number	If the items are in a book or a scrap book and that scrap book has a section this is where that is recoreded
Page#	If a scrap book has a set of pages then this is where they are recoreded
Duplicates	This is where the Photo ID of a duplicate item is referenced.
Old Inventory Number(s)	This is the inventory number of an item if it were part of another invenotry system
Photographer	This is the name of the photographer
Subject 1 (who)	Who is in the photo, this should be an unlimited field. That is sveral names should be able to be added to this.
Subject 2		Who is in the photo, this should be an unlimited field. That is sveral names should be able to be added to this.
Subject 3		Who is in the photo, this should be an unlimited field. That is sveral names should be able to be added to this.
Subject 4		Who is in the photo, this should be an unlimited field. That is sveral names should be able to be added to this.
Subject 5		Who is in the photo, this should be an unlimited field. That is sveral names should be able to be added to this.
People group	This is the name of the people group meneined in the ISO 639-3 codes
ISO 639-3 Code	This is the ISO 639-3 code of the people group being photographed
When was the photo Taken?	The date the photo was taken
Country	The country in which the photo was taken
District/City	This is the City where the photo was taken
Exact Place	The exact place name where the photo was taken
What is in the Photo (what)	This is an item in the photo
What is in the Photo	Additional what is in the photo
What is in the Photo	Addtional what is in the photo
Why was the Photo Taken?		This is to help metadata providers think about how events get communicated
Description	This is a description of the photo’s contents	This is not a caption but could be used as a caption
Who Provided This Meta-Data? And when?		We need to keep track of who is the source of certain metadata to understand its authority
Who Provided This Meta-Data? And when?		We need to keep track of who is the source of certain metadata to understand its authority
Who Provided This Meta-Data? And when?		We need to keep track of who is the source of certain metadata to understand its authority
Who Provided This Meta-Data? And when?		We need to keep track of who is the source of certain metadata to understand its authority
Who Provided This Meta-Data? And when?		We need to keep track of who is the source of certain metadata to understand its authority
I am in this photo and I approve it to be on the internet. Put in "yes" or "No" and write your name in the next column.	Permission to distribute
Name:	Name of the person releasing the photo
How was this photo digitized?	Method of digitization and the tools used in digitization
Who digitized This photo	This is the name of the person who did the digitization