The new indexing program is still a work in progress, said Flinders. He gave us a walk through like you do with a home during construction. Imagine a program where partners (like archives and societies) create and manage their own indexing projects. Imagine users being able to share an image with other indexers to ask opinions on what is written. And the system knows when you’ve got enough experience to serve as an arbitrator and invites you to become one. Imagine a user who finds a record on FamilySearch.org being able to send a thank you note to the indexers. I wasn’t certain if Flinders was saying that the new program would be able to do all these things, or if he was just getting our creative juices flowing.
The new program will have a number of key improvements according to Flinders.
- It is a browser based tool.
- It is part of the Familysearch.org website.
- It has mobile device support.
- You can join multiple groups.
- For members of the Church of Jesus Christ of Latter-day Saints, it knows what ward (congregation) and stake you are in, even if you move.
- It supports achievements and recognition.
- It has social media integration, primarily Facebook, but others will be introduced. This will allow you to, for example, see who else is an indexer.
- It allows external project administrators. That is, societies and archives can administer their own projects.
- It allows multiple values per field. Today, the A keyer, B keyer, and arbitrator could all have specified different values, but only the arbitrator’s is saved. The new system allows all three opinions to be saved, and when searching, any of the values will match. [This was met with applause.]
- It persists highlights to be used later on. When you do a search, wouldn’t it be nice to have highlights on the image? Eventually that will be part of the search experience.
- New volunteer training is integrated into the real indexing tool. Today a replica is used that doesn’t completely replicate the real tool. Real record images with known values are used.
- It has improved training and help.
- It has enhanced project selection.
- It has improved workflow and quality. With today’s system, every batch is indexed twice, and then differences are reviewed by an arbitrator. FamilySearch has a new workflow that they believe will produce equal quality, but will produce more records for the same amount of work.
- It has better reports for administrators and group leaders.
- It uses cloud technology which allows it to scale better to handle more users.
- It has performance testing. (I don’t recall what that means.)
- It has a formal API. I think that means other companies can write their own indexing programs that utilize the guts of FamilySearch indexing. But I didn’t ask.
Some of the new features are already live, said Flinders. On FamilySearch.org click on Indexing. There you can see real-time statistics and enhanced tools for finding an indexing project.
- Sample images will be from the project. Today, sample images might be from a different project!
- While it won’t be available for all projects, sometimes you’ll be able to select batches from a location within a project. For example, you could pick the county from a project to index records of a state.
- Today, there is a choice of form or table data entry. In the new system, the form entry sub-window can be moved around. The new system also has other entry modes: column entry, row entry, and Inline. I’m afraid I didn’t notice these well enough to describe how they work or appear.
- A filmstrip of images gives access to more than just the next and previous images.
- After submitting a batch, the system will recommend batches for you to do next, including some from the same project, from a new project, or from those you’ve checked out.
The new program has a new workflow. I didn’t catch if it is a replacement, or an alternative to the current A/B/arbitrate flow. The new flow is called “A + Review.” One person indexes. A reviewer checks all fields, looking for and correcting errors. If they change less than a certain number of fields, it is allowed to go on, otherwise, it goes back to another indexer. The new flow is at least 30% more efficient, said Flinders.
The new program adds a My Indexing page. This is a replacement for the first screen of the current indexing app. It lists your open batches and their percentage completion. It has the system messages. Scroll down to see a progress graph, a list of groups that you belong to, and an activity feed. The feed looked a bit a Facebook or Twitter list of items, happenings, accomplishments, and the like.
Each group has its own home page, which also struck me to be a bit like Facebook pages. It would have a group progress graph, a list of group members, and I think their roles. (The term arbitrator has been replaced with reviewer.)
The system will have new roles. (indexer and arbitrator are example roles from the current system.) New roles are setup and linking. Those doing project setup identify non-indexable images and divide images into batches. Linking matches existing indexes to images.
Flinders told us the roadmap for the project. An alpha version of the core indexing experience is done. During the first half of the year they are finishing up the features shown today. (I might add a reminder that this was a “house under construction walk through.” I always take these feature lists with a degree of skepticism.) The current target is to have the new program in the second half of the year, perhaps August. There will be a period of time during which the two systems will coexist during project completion and indexer transition.
In a move to give subtle preparation to the appearance of the new system, FamilySearch will introduce a change at the end of February to the current indexing system. It will get a really minor facelift, including a new font in titles, the new FamilySearch logo, a new green bar with batch name, and other fine changes that are in line with the new indexing system.