As we are about to launch new metadata crowdsourcing tools to Historypin's Year of the Bay project, I would like to briefly have a look at one of ways Historypin has focused on user interaction research leading up to this exciting launch. For the past few months, we’ve started crowdsourcing Year of the Bay photo mysteries. Each week, we’ve been picking photos pinned to the project that we need more information on (ones we've either stumbled upon or ones contributed by users and partner institutions) and release them over social media, our chosen platform. Asking for the Historypin mainstays of location and dates if they are needed, we also highlight photos that need business information, specific event information, or any other aspects of a photo that can lead to larger conversations. So far, it has been a great chance for us to harness the power of crowdsourcing to dig deeper into local history, while at the same time helping to build more accurate data for individual and institutional collections.
How did we want people to interact with this content? For this "history mystery" series we have chosen to drive traffic mainly over social media, as this most recent phase of development is about monitoring interaction over various platforms to inform the development of better Historypin crowdsourcing tools. Posting the photos to our Facebook and Twitter feeds with the hashtag #historymystery, we record not only the comments made but also any analytics to see the kinds of photos users are interacting with.
 |
| Our first #historymystery: Photo of an unidentified submarine, contributed by the San Francisco Public Library. We were able to narrow down the date, find the submarine's origin, and lined up a few suggested models of subs. |
Basic data collection:
For this experiment, we are managing data over four social media spaces: Historypin's Facebook and Twitter feeds, and Year of the Bay's Facebook and Twitter feeds. To avoid double-posting, we chose to create mystery posts as Historypin, and then "share" them on Year of the Bay's social media channels.
Over these four spaces, I harvest data on a spreadsheet detailing some basic information obtained for each mystery: date posted, title, links to the original Twitter and Facebook posts, how many retweets, shares, comments, who I specifically targeted to help solve the mystery (if any), etc. Also, because we chose not to drive conversations over Historypin itself, I manually pick out the most useful comments that we have gotten via social media to add as "comments" to the original pin on Historypin, as well as including a link to the original Facebook mystery discussion post for reference. I always make sure to connect with the original pinner as well, to let them know about the discussion happening around their photo. To make such a task more manageable, we release two mysteries per week instead of every day.
 |
| After releasing a #historymystery photo over social media, we add the original Facebook discussion thread along with any updated information to the pin on Historypin. |
What has worked best so far:
This crowdsourcing experiment has been a great way to reach out to specific audiences. For instance, we may reach out to local transportation enthusiasts using Twitter to drive conversations around a particular car model within a photo, or work with a neighborhood history group on Facebook to help identify a location they might know something about. Whether calling out to someone directly for help in a mystery post or sending a mystery to them, letting the local community know that we are interested in their expertise and feedback has provided many enriching conversations and interactions around historical content.
Also, targeting specific individuals and groups has often helped to spread local history conversations farther over social media. For example, asking a Facebook page such as Lost San Francisco for help has in the past lead to their sharing of our mystery posts with its thousands of followers. As Historypin is a global entity, being able to tap into a community specifically geared toward local history such as Lost San Francisco almost always proves fruitful in terms of user interaction.
 |
| One of our #historymystery posts "shared" by Lost San Francisco; as result, the photo received over 7,000 views and 50+ shares on Facebook. |
Also, thanking individuals is especially good practice, to let our users know that anyone can contribute and aid in enriching content. A very important aspect of commenting that we aim to encourage is the citing of sources; to that end, we try to tweet about and highlight the practice of users who add comments or suggestions with good citations, such as this incredible map below:
 |
| A map made by user Chris Salvano in order to help us solve a 1906 earthquake #historymystery, one of the best examples of user citations we have seen throughout this series. |
What are the problems with this kind of crowdsourcing?
Although we are harvesting data within what appears to be four manageable social media spaces, once a Facebook user "shares" our photo post with their community then this becomes its own post that requires monitoring. For example, if the California Historical Society shares our mystery post, then this becomes a separate post on their wall with its own set of comments, "likes," and shares. Historypin will not receive any notifications for activity there, but instead I will have to manually check it under the "Share" button on the original post.
Also, "likes," shares, and retweets are not always indicative of how people are interacting with these mysteries. For example, users might want to comment anonymously, and thus may not comment at all for fear of being refuted on a public thread. There also may be pressure to only comment with an "answer" rather than any other kind of observation.
From the managing side, if choosing to keep up with the lightning-fast pace of social media and responding instantaneously, monitoring comments can also become very time consuming if not kept in check.
Overall observations so far:
One of the best aspects of this experiment is being able to communicate more closely with our user community, both on a local and global level. Releasing the photo mysteries over social media, where communication is instantaneous, produces many "in the moment" conversations in which one comment may spur on an entire thread. These small history conversations, whether a historical factoid about the photo subject or a personal memory triggered when viewing an image, are arguably more important than finding an "answer" to a mystery.
And though "Likes," "Shares," "Retweets," etc. can only tell us so much about the way our users interact with these photo mysteries, having local history become a regular part of the social media experience is something that this kind of crowdsourcing helps to make commonplace. Local history is also personal history; as the series has progressed, we have gotten more user-submitted mysteries which means that we are able to release a mixture of personal and institutional content for discussion. Ultimately, getting out the idea that solving a local mystery on a global platform is possible, that #historymystery solving can be a social experience, is one of the most powerful ideas behind crowdsourcing for the digital humanities and what Historypin is all about. With the launch of the new metadata crowdsourcing tools, we envision Historypin becoming a hub for these types of conversations, with mysteries solved along the way.
-Kerri Young, Historypin Project Officer
