Building the social and technical bridges to enable open sharing and re-use of data
RDA EURDA USCONTACT US LOGINREGISTRATION

Searching for connections: Data search, web search and literature search

You are here

Home » Plenaries » Searching for connections: Data search, web search and literature search
24 Feb 2022
Group(s) submitting the application: 
Data Discovery Paradigms IG
Meeting objectives: 

  • To update the group progress

  • To update on latest studies of data search 

  • To identify a work plan towards a best practices output for the topic

Meeting agenda: 

Collaborative session notes: https://docs.google.com/document/d/1wLRiA36iLJ97SlTCk1ohNswSFyYc6g0roL-b...

  1. Introduction of the group (5 minutes) - Kathleen Gregory

  2. Four flash presentations of invited speakers on the topic of data search studies (12' each : ~50 minutes). The goal of the presentations is to identify connections/gaps in existing work. We sent an email soliciting presentations on 21 Feb, with the aim of recruiting new perspectives.  This list includes confirmed speakers. 

  • Brigitte Mathiak: Web Search and Literature Search During Data Discovery
  • Tim Suber: A User Model for User-Adaptive Searching in Data Portals
    • Abstract: Data portals host a large amount of data sets. Finding the right data set can be challenging. The goal of this work is to guide the user to the correct data set. This is implemented by creating a user model to make the search user-adaptive. This work focuses on supporting the user in using search filters by predicting future interactions. For this purpose, a model of partially-ordered sequential rules is trained based on previous user sessions. The model was tested on an existing data set, containing records of a geospatial portal. It was able to predict subsequently chosen search filters with an accuracy of 51%, which is a significant improvement over a random prediction under the same conditions with 11%. The results show that the development of a user model is a suitable process to predict search filters and can improve user experience in data portals. 
  • Graham Parton: Levelling up data discovery at a UK domain repository: How the UK’s CEDA Archive uses a range of tools to aid users discover what they need… and the gaps that still remain.
  • Mingfang Wu: Overall user study design for a national data discovery portal / log survey or interview analysis from a national portal

  1. Discussion on the following topic statement (to be adjusted based on the discussion) (20 minutes)

  • Topic statement: How can we enhance data discovery through the connection/integration of multiple resources, e.g.  web search,  literature search, data repository search, and what would be the corresponding best practices / recommendations that could be offered to repositories, aiming for user studies?

  • Outcomes:

    • Scope and goals

    • Roadmap towards an output

       4. Next steps and wrap up (10 minutes)

 

Target Audience: 

  • Researchers who conduct user studies for understanding more about an individual’s data discovery process 

  • Data managers/providers who are responsible in describing data and making data findable

  • Data managers to investigate whether any user studies have been performed at their site

  • Attendees with some prior preparation or insights for their institutional / personal data discovery approaches would benefit more during the group discussion parts of the session.

Group chair serving as contact person: 
Kathleen Gregory
Brief introduction describing the activities and scope of the group: 

The objective of this IG is to provide a forum where representatives from across the spectrum of stakeholders and roles pertaining to data discovery can work together to identify, study and make recommendations concerning issues related to improving data discovery. The goal is to produce concrete deliverables that will be recognised and valued by the research and data communities.

This group was officially endorsed at RDA P9. The group has worked on the following task forces, namely:

  1. User studies in data discovery (ongoing)

  2. Data/Metadata granularity (ongoing, a BoF has been submitted)

  3. Using schema.org for research dataset discovery (This task force has spun off to the Research Metadata Schemas Working Group, which was endorsed in Sept. 2019).

  4. Initial four task forces from the group:

    1. Relevancy ranking (completed)
       

    2. Use cases, prototyping tools and test collections (completed)
       

    3. Best practice for making data findable (completed)
       

    4. Metadata enrichment (closed)

Short Group Status: 

The DDPIG has been established and endorsed as an IG during P9. The group started with four task forces around target data discovery topics soon after P9. All task forces actively explored their topics, and reported progress and outputs at consequent plenaries. At P11, the first three task forces were officially closed, and a discussion on new Task Forces took place, focusing during P12 primarily on Schema.org and Data Granularity. After P13, a case statement for a Research Schemas WG was submitted, the case statement was endorsed in Sept. 2019, just before P14.

Type of Meeting: 
Informative meeting
Additional links to informative material: 

The group has delivered the following three supporting outputs: 

Slides from previous plenary sessions: 

Slides from earlier plenaries are available from the group page.

Avoid conflict with the following group (1): 
Research Metadata Schemas WG
Avoid conflict with the following group (2): 
FAIR for Research Software (FAIR4RS) WG
Meeting presenters: 
DDPIG Co-chairs + Volunteers for presenting lightning talks or leading the discussion
  • 891 reads