A New National Access Point (NAP) End of Alpha Show, Tell & Ask – 22 May 2020 Moving Britain Ahead May 2020 Official A New UK National Access Point 1
Hello 2
Housekeeping and recommended etiquette This deck will be shared with you after today ● ● Please stay on mute throughout ● You are welcome to pose questions in the chat box at any time ● There will be a Q&A towards the end of the session, when we will try to respond to questions as much as time allows 3
What is this session about? 1 2 3 4 5 Refresher: Users Prototype Beta Q&A What did we demo and recommenda- talk about at Research tion the previous findings Show & Tells? 4
Proposition A National Access Point that improves transport data discoverability to improve data visibility and reuse across different sectors. Improved search to make experience of ● ● More useful metadata, use cases and data finding data better, easier and quicker visualization to improve users ability to find the data they need ● A central place that signposts to both public ● Usage information from data users’ feedback and commercial data across the UK to provide and engagement to give insight into how data is a complete view of data sources used and how it could be improved by publishers 5
Refresher We established that Alpha test assumptions recorded as hypotheses; it focuses on learning, not delivering a finished product We shared the four hypotheses on 1) metadata, 2) metadata submission, 3) discoverability, and 4) feedback and engagement We introduced personas Private sector Public sector We talked about our approach to prototyping: Mock-ups + CKAN platform Rob Claudia Ben Elena Traffic Manager at Logistics at delivery Director at private Data Analyst at DfT Local Authority company data consultancy 6
Coronavirus response It took nearly 2 weeks of time to identify relevant datasets for coronavirus response. The response team believed this time could have been massively reduced had the NAP existed. Source: ● Stakeholders involved in the building of the national dataset of LA live UTC data as part of covid response 7
Users 8
Users The core users of the NAP are those in roles and Across sectors ● Central Government teams that use and publish transport data. ● Local Authorities ● Commercial sector Including: ○ Data consultancy ○ Data companies that create, use, provide data ○ Those that offer and create products and services Data Analysts ● that use transport data or create data via operations ● Data Architects ● Research Traffic Managers ● ● Logistics, parking, taxi, road freight teams Smart city programme teams ● ● Technologists Data scientists ● ● Data consultants Waste service teams ● ● Software developers 9
Data publisher user mapping Understands and advocates for data (I publish because I think it’s important to) Needs standards Academic researcher - P2 and API’s Potential users of NAP have different motivations and different levels of data literacy. Mapped here are some of the data publishers we spoke with during Alpha. Smart cities programme manager, Local Authority - P3 High technical expertise Low technical expertise Data architect, Central Data management specialist, (I think in data models) (I think in spreadsheets) Government - P7 Public data company - P10 Commercial data service company director - P5 Traffic Manager, Performance manager, Central Local Authorities Government agency - P1 Needs ease of managing data Doesn’t understand or advocate for data (I publish because I’m required to) 10
The general public is not in our target user group. Although the general public will have access to a NAP, we do not expect them to be users of service. Unlike the ONS service, the NAP will not provide pre-analysed data and insights. 11
Prototypes 12
Prototypes created and used for testing CKAN platform Mock-ups For things we decided we can’t prototype We used a mix of clickable/static to test effectively with just mock-ups, we digital mock-ups for certain parts of decided to use CKAN, an open source the design. data portal, and modify it as much as we could for testing purposes. 13
Alpha prototypes overview and demo To test the prototypes we created test tasks for users to complete around road transport data. For data users: For data publishers: Searching for data Understanding if and how publishers would submit ￮ CKAN instance start from home page (prototype not in the open domain; metadata see the next three slides for screen captures) Metadata Public data mock-up ￮ Task: “Find which companies are currently officially registered to run bus ￮ services in the NE of England. Find it in a format you could use in excel.” How usage insights provide value ￮ Task: “Can you find road safety statistics for road accidents for 2018” Analytics page mock-ups for data publisher ￮ Same Understanding and assessing suitability of data through metadata prototype Metadata Public data mock-up ￮ ￮ Task: “ Ok, suppose you found some data regarding road casualties, you click to view details and this is what you see.” ￮ Task: Find information on the page: licensing, formats, missing information, data coverage map Testing what commercial data/metadata could look like ￮ Metadata Commercial mock-up Giving feedback to data owners/publishers Feedback page mock-ups ￮ 14
CKAN prototype - discoverability (search) The starting point for many users when landing on the front page of the NAP is the search box. This allows free-form basic queries, as well as structured logical queries. The road transport specific ‘topic’ labels are shown prominently here to prompt users by a different search facet. 15
CKAN prototype - discoverability (browse) The wider landing page provides alternative ways to search and navigate to the desired dataset, including by navigating by transportation mode. 16
CKAN prototype - discoverability (results page) If a user searches or browses by transportation mode, publisher, other category, or by using the featured dataset boxes, they view a result set. This summaries here help give a user a quick insight into the nature of the dataset and the format labels make it easy to know in what form the dataset is available. 17
High-level findings ● Good quality, well maintained metadata does make it easier for users to judge the suitability of a dataset for their task. ● However, a system with good data visualisation, previewing, searching and browsing functionality is needed to take advantage of this metadata and improve the user experience. ● Data publishers are able to provide the required metadata but some will need additional support and guidance in the submission process and preparing data. ● Commercial data publishers have different needs compared to public sector data publishers. They perceive the NAP as a potential marketplace for selling services rather than sharing data. Data analytics would be used by data publishers to understand what data was useful for ● the community; it would also help them justify the resources they need to publish and maintain data. ● Data user feedback was seen as valuable to most participants. Some had concerns about the feedback being public and preferred a private solution; some acknowledge existing open forums for discussion and advocate a community of interest. 18
Discoverability versus quality metadata The initial focus of the NAP was to provide high quality metadata in order to improve discoverability. Our research has shown that metadata is not the only consideration in discoverability. In addition, there is often a trade-off between improving overall discoverability and improving metadata. High quality metadata Lower quality metadata ● Increased workload for data publishers ● Reduced workload for data publishers Improved clarity for data users Increased chance of submission to the NAP ● ● ● Lower chance of submission to the NAP ● Improved discoverability for data users ● Reduced numbers of datasets represented, through increased number of datasets reduced overall discoverability for data users 19
Beta recommendation 20
Overall focus The Beta should focus on discoverability and curation . The NAP should take a minimum-viable approach to metadata, ensuring the basics are captured. This approach recognises that as the effort to add data to the NAP increases, the likelihood that a wide range of data sources will be entered reduces. Improving and deepening metadata should be a long-term goal of the platform. Multiple data producers have told us they would take a lead from the NAP on metadata standards. However, the immediate opportunity (as shown by Covid-19) is to make available data sources more visible and accessible. In the absence of metadata, data previews have been shown to be valuable in users decision making. The Beta should be run for a minimum of 6 months to understand how data is maintained and changes over time. 21
Recommend
More recommend
