⬆ Design Documents Overview

DocNow Conversation

creator: Ed Summers
created: 2019-05-21


Midway through Phase 1 of the project we realized that we couldn’t simply build an easier to use, more accessible, or faster Twitter data collection tool and consider our work done. There were several reasons for this shift:

Hearing from activists about the many ways they wanted their content to be used as part of a record of the Ferguson protests. Hearing from a contingent of archivists who thought that content creator’s consent was integral to archiving social media. Acknowledging that Twitter data is just a thin veneer over the rich documentary records that are available for an event like the Ferguson protests. Learning about the options that already exist for institutions to bulk collect large amounts of social media content, either their own (tools like Archive-It, ArchiveSocial) or others (GeoFeedia, SocialFeedManager, etc)

We recognized that there is a need for tools that allow archivists to center the agency of content creators on Twitter to allow them to meaningfully participate in the creation of archives. But how can we do this in the DocNow application, especially when you consider that there are potentially thousands of users tweeting with a particular hashtag like #ferguson, and there is no real means of communicating with them short of resorting to what Twitter would regard as unsolicited spam?


The DocNow application has the following requirements to address this need:

These items are out of scope, or non-requirements:


To better understand these requirements we conducted a literature review of the Human Computer Interaction (HCI) literature to examine how interface designers and information technology researchers are thinking about and responding to the role of consent in data collection activities. You can read a write up of some of the findings from that literature review at https://news.docnow.io/designing-for-consent-2f9e9cb2ab4f.

There was wide agreement that consent, as it has been traditionally conceived, is broken on the web. The primary cause for this breakage is that it is much more complicated to determine a priori the many contexts that our personal data can travel through, and that people have different expectations of consent in these varied contexts (Nissenbaum, 2011). There is a set of design factors that can be used to more effectively empower users to make decisions about their privacy (Acquisti, 2017) but these same factors are being used to deprive users of agency around their data.

One useful way for thinking about ways we can build a conversation around consent into the DocNow application is provided by the Human Data Interaction framework (Hutton & Henderson, 2017) which emphasizes three broad areas of concern:

The goal is to design into the DocNow application a user experience for both the archivist and content creators that speaks to these three concerns.

Action Items