Interfacing Session

Tuesday 7 June 2011 afternoon

 

This session addresses questions like: Are current search problems in the patent domain caused by the fact that the search interfaces have been designed for other domain searches? How could the perfect interface for patent search look like? The session will show new developments in the field including the demo of an interactive interface with higher level search functionalities that allows for direct manipulation of queries, result sets and documents.

Schedule:

  • Session Keynote: Tony Russell-Rose, UXLabs
  • Demos:  Interactive patent search interface
  • Q&As

Keynote

Designing the Search Experience

Tony Russell-Rose, UXLabs

The landscape of the search industry is undergoing fundamental change. In particular, there is a growing realisation that the true value of search is best realised by embedding it a wider discovery context, so that in addition to facilitating basic lookup tasks such as known-item search and fact retrieval, support is also provided for more complex exploratory tasks such as comparison, aggregation, analysis, synthesis, evaluation, and so on. Clearly, for these sorts of activity a much richer kind of interaction or dialogue between system and end user is required. This talk examines what forms this interactivity might take and discusses a number of principles and approaches for designing effective search and discovery experiences.

Tony Russell-Rose currently is director of UXLabs, a consultancy specialising in user experience research, design and analytics. Before founding UXLabs, Tony was Manager of User Experience at Endeca, a software company specialising in innovative solutions for information search and business intelligence. Prior to this he was technical lead at Reuters, specialising in advanced user interfaces for information access and search. And before Reuters he was R&D group manager at Canon Research Centre Europe, where he led a team developing next generation information access products and services. Earlier professional experience includes a Royal Academy of Engineering fellowship at HP Labs and a Short-term Research Fellowship at BT Labs.

His academic qualifications include a PhD in human-computer interaction, an MSc in cognitive psychology and a first degree in engineering, majoring in human factors. Tony also holds the position of Honorary Visiting Fellow at the Centre for Interactive Systems Research, City University, London.

He currently acts as vice-chair of the BCS Information Retrieval group and chair of the IEHF Human-Computer Interaction group.

Interfacing Demonstrations

Demo: Using ezDL as a Search Interface for Patent Searching

Matthias Jordan, Universität Duisburg-Essen

Patent searching is an interactive and complex endeavour. Support for iterative query sessions, for search history management, and for organizing and working with results sets is needed by patent search professionals.
 
Daffodil was created as a multi-agent search system for heterogeneous digital libraries, based on models proposed by Marcia Bates. It provides higher-level search functionalities in form of loosely coupled tools that implement so-called stratagems. Daffodil also supports the berry-picking model of information searching. EzDL is a re-implementation and partial re-imagining of the Daffodil software developed at the University of Duisburg-Essen. It simultaneously searches several digital libraries mainly from the area of computer science, and is continuously expanded with new interactive functionalities. While command line interfaces are not uncommon even among modern patent search solutions (eg. STN Express), the ezDL UI allows for direct manipulation of queries, result sets and documents.
 
We propose to demonstrate the suitability of the ezDL software suite as a single user interface for patent searching. For this, we intend to index the MAREC XML data set and make it searchable through the ezDL search tool, which will be extended to support many of the query formulation functionalities important to the patent community. The ezDL session support will allow to save and continue search sessions with full activity history. This is an important feature for patent search professionals who need to account for actions taken and results examined to provide evidence that they performed an exhaustive search.
 
In addition the system will be extended such that previous queries can not only be browsed, filtered and re-used, but also combined into new queries. Searchers will be able to sort, filter and group results by several criteria, store and exported them, and extract common terms, concepts, authors, or sources from result sets. The result documents themselves will be viewable within ezDL, and be enriched by keyword highlighting and the possibility to use them as a springboard for new search queries.
 
We are looking forward to demonstrating the utility of our approach and the usability and learnability improvements over traditional interfaces using exemplary patent search tasks.
 

 

Matthias Jordan

Matthias received his diploma in computer science from the University of Dortmund, Germany, in 2005. He worked as code monkey for Framfab in Hamburg, where he helped develop and maintain web sites for Postbank and the International Olympic Committee. In 2008 he joined Norbert Fuhr's research group where he teaches information retrieval and databases, develops ezDL and pursues a Ph.D. in interactive IR.

 Sponsored by

 

 In cooperation with

 

 Supported and endorsed by


WON

CEPIUG