Hivemind Release - Document and text annotation, instance priorities, and more..
We are always thinking of new and exciting ways in which Hivemind can facilitate the data collection that our clients perform. With this release, we are delighted to announce that Hivemind now supports both PDF and text annotation in Workbench forms.
The new tooling in Workbench allows the contributor to navigate through one or more PDF files, drawing boxes around relevant sections of the document and marking these against a configurable list of classifications, with the captured sections being displayed in a separate panel for easy reference. The results are then provided using the usual mechanisms in a specific format denoting the location of each annotated section, giving both page number and position within the page.
Our text annotation component likewise allows you to create a task using text snippets, and have your contributors annotate and tag this text against a configured set of categories. Our intuitive tooling provides a quick and easy way to highlight and classify words using simple mouse clicks, whilst more detailed snippets of text such as partial words and individual characters can be tagged using a precision function. The position, classification and value of the selected snippets are then provided in the task results.
New ways to orchestrate instances
We’ve introduced a new way for you to control the order instances are completed in this release. By specifying the priority on the instance, you can dictate the order in which instances are presented to contributors. This is particularly useful in cases where you need instances on a task done before others, for example, in cases where they’re needed to progress time-sensitive workflows.
The task creation wizard has been updated to map an integer column in your source CSV file directly to this new parameter. A higher numerical value indicates a higher priority and will be presented to a contributor before lower priority instances. You can also choose to omit the value entirely, in which case the instance will be treated as if a 0 was passed in, the default priority. Using this mechanism, you can also mark instances as low priority by passing in a negative number.
This release also introduces a new internal method to the way instances are selected to be worked on within priority groups or when no priority is specified. In the past, you may have found that the progress of a task was slow when work initially started due to the way Hivemind emphasised independence of the responses. With this change, instances will be presented to contributors in a way that aims to deliver a steadier rate of completion of instances throughout the life of the task, while maintaining the principles of independence between contributors who happen to be working on the same instances in near succession. The biggest benefit of this new method is a reduction in the time before the first results are generated, allowing you to gauge the suitability of your data quality parameters and overall task design.
Further changes to the form builder
This release also introduces further enhancements to the form builder, specifically how enumerated fields are handled. You can now define the enum and enumNames parameters on a dropdown to gain a finer balance between ease of entry for the contributor and easier processing of pre-mapped results.
