Exploratory Search Interface for Knowledge Databases

Information

• Publication Type: Bachelor Thesis
• Workgroup(s)/Project(s):
  • Visualization Group
• Date: December 2021
• Date (Start): April 2020
• Date (End): December 2021
• Matrikelnummer: 01402617
• First Supervisor: Manuela Waldner

Abstract

Visualization has always been a powerful tool to effectively convey knowledge and information. It has also gained attention in the context of search, in particular for newer visualization techniques like “Multifaceted Search” and “Exploratory Search”. There are currently many tools and websites that still rely on an explicit search function or an alphabetically ordered glossary of terms to allow users to filter and browse resources. This results in many useful resources not being discovered by users because of a lack of proper search tools. Exploratory Search is more open-ended, allowing users to search even if they do not exactly know what they are looking for.

This thesis proposes an adaptable, modular, web-based prototype of an exploratory search interface. The goal of the prototype is to serve as a basis for the evaluation of exploratory search interfaces for a wide variety of use cases. In contrast to many existing Exploratory Search tools, this prototype does not require rich meta-data to be present in a dataset. By utilizing an optional preprocessing step to extract named entities via Natural Language Processing, the prototype is compatible with most text-based datasets. The search interface consists of a word cloud created by a force-directed layout algorithm that places related entities close to each other. This interface also serves as the main filtering option, which keeps the users’ focus on the word cloud. After selecting interesting entities, matching documents can be browsed in a list view.

Additional Files and Images

Weblinks

No further information available.

BibTeX

@bachelorsthesis{bugnar-esk-2021,
  title =      "Exploratory Search Interface for Knowledge Databases",
  author =     "Philip Bugnar",
  year =       "2021",
  abstract =   "Visualization has always been a powerful tool to effectively
               convey knowledge and information. It has also gained
               attention in the context of search, in particular for newer
               visualization techniques like “Multifaceted Search” and
               “Exploratory Search”. There are currently many tools and
               websites that still rely on an explicit search function or
               an alphabetically ordered glossary of terms to allow users
               to filter and browse resources. This results in many useful
               resources not being discovered by users because of a lack of
               proper search tools. Exploratory Search is more open-ended,
               allowing users to search even if they do not exactly know
               what they are looking for.  This thesis proposes an
               adaptable, modular, web-based prototype of an exploratory
               search interface. The goal of the prototype is to serve as a
               basis for the evaluation of exploratory search interfaces
               for a wide variety of use cases. In contrast to many
               existing Exploratory Search tools, this prototype does not
               require rich meta-data to be present in a dataset. By
               utilizing an optional preprocessing step to extract named
               entities via Natural Language Processing, the prototype is
               compatible with most text-based datasets. The search
               interface consists of a word cloud created by a
               force-directed layout algorithm that places related entities
               close to each other. This interface also serves as the main
               filtering option, which keeps the users’ focus on the word
               cloud. After selecting interesting entities, matching
               documents can be browsed in a list view.",
  month =      dec,
  address =    "Favoritenstrasse 9-11/E193-02, A-1040 Vienna, Austria",
  school =     "Research Unit of Computer Graphics, Institute of Visual
               Computing and Human-Centered Technology, Faculty of
               Informatics, TU Wien ",
  URL =        "https://www.cg.tuwien.ac.at/research/publications/2021/bugnar-esk-2021/",
}