Information Cognition (IC): Enabling the Human Cog in the Information Retrieval Machine

Project: Research

Description

"SUMMARY

This research programme will investigate the feasibility of using visual representations for the secure navigation and search of large, complex, multimedia data sets. It draws upon prior research that shows the human visual system has a powerful ability to recognise and classify objects in 3D environments. To do this the project will create an experimental platform that combines human cognitive abilities with recent advances in design information management, computer graphics and data mining. The resulting system will support the systematic study of how advanced visual interfaces impact on a user's ability to both find individual items and identify patterns, or oddities, within subsets of the data.

The veracity of the project's experimental methodology is ensured through close collaboration with an international research centre supported by a consortium of leading defence contractors (ie Rolls Royce, Boeing etc). This partnership will ensure the project is focused on the challenges faced by the defence industry and can assess its outputs using a realistic bench mark environment."

Key findings

This work examined the human ‘cog’ within a data visualization system; testing the hypothesis that human beings find the recall and recognition of 2D and 3D shapes and environments so intuitive and effortless that any system for the effective retrieval and use of data should make use of this fact. This first development phase has indicated that the SIZL system does help users search for and identify relationships between documents in large datasets, when compared to a traditional text-based system. A range of quantitative and qualitative analysis methods were used and the key findings are as follows:

The KEY FINDINGS from ANALYSIS of the TIME and ACCURACY of the participants’ responses were:
• There does not seem to be a correlation between people’s ability in File Manager and their performance with SIZL.
• When data retrieval is the task, File Manager is much faster.
• When the task is about making relationships between files, SIZL can achieve the same accuracy as File Manager in less time – almost half the time.
• SIZL outperformed File Manager on more complex questions, i.e. responses requiring relationships between files and the information.
• When using File Manager participants were getting more right relationships as time went on, however in SIZL the opposite was true. Participants who spent the longest time in SIZL tended to ‘get lost’ and their accuracy decreased with time.
• People were most enthusiastic about SIZL’s multi search facility, and generally preferred the visual layout.
• The participants’ most collective complaint was errors in using the system.

The KEY FINDINGS from the PARTICIPANTS' SURVEY were:
• Participants responded highly positively to the multi-search function.
• The majority of participants expressed a preference for the visual layout of the SIZL system.
• Some participants struggled to recover when errors occurred in the system.

The KEY FINDINGS from PARTICIPANT OBSERVATION and NOTE TAKING were:
• The multi-search function was the most useful and most popular feature of the SIZL system.
• Participants responded positively to the visualization provided by SIZL.
• Participants used the SIZL system to navigate their own problem-solving processes, i.e. the human 'cog’.
• The zoom and timeline functionalities were not used as frequently as expected due to lack of highlighting its importance.

The KEY FINDINGS from the FOCUS GROUP were:
• In correlation with the qualitative data findings, the multi-search function was the most popular feature of the SIZL system.
• The timeline and zoom features were not used readily. This was due partly to not making this feature explicit in its introduction.
• Also mirroring the key findings from the participants’ tests, errors in the use of the SIZL software were highlighted as a key issue, but not so much as to affect functionality in testing.

Overall, across all forms of evaluation, the multi-search function proved extremely useful to participants in searching and browsing for information. In addition, participants responded positively to the visualization provided by SIZL.
StatusFinished
Effective start/end date30/05/1231/08/13

Funding

  • EPSRC (Engineering and Physical Sciences Research Council): £117,053.00

Fingerprint

Information retrieval
Managers
Visualization
Data visualization
Testing
Computer graphics
Contractors
Information management
Data mining
Navigation
Industry