Information

Abstract

The general functions of current web search engines are well established. A box is provided in which to type the queries and the engine returns a result list which users can evaluate. The autocomplete suggestions assist users in defining their problems, however there is a lack of support for an iterative manual refinement of the query. This additional aid can be beneficial when users not know the exact terms to describe the concept they are looking for. Therefore, the goal of this project is to show searchers how a slight variation of the query changes the results. With this information, they then can perform a targeted refinement of the query to access useful information sources. To achieve this goal, each part of the searcher’s query is varied with a thesaurus that provides synonyms for the individual query terms. While performing the user’s original query in a normal fashion, variations of this query are conducted in the background. To provide a concise visual summary of the query results, text mining techniques are performed on all gathered results to retrieve the most important key terms for each query variation. This procedure results in a visual overview of what the searcher’s query finds together with what could be found with a slight variation of the query. Additional queries should make users aware that alternative queries may be more appropriate when their original query is poorly formulated. In conjunction with some interaction tools, the goal is to reduce the burden of refining search queries and therefore making searching the web less complex.

Additional Files and Images

Additional images and videos

Additional files

Weblinks

No further information available.

BibTeX

@studentproject{mazurek-2017-vows,
  title =      "Visualization of Thesaurus-Based Web Search",
  author =     "Michael Mazurek",
  year =       "2017",
  abstract =   "The general functions of current web search engines are well
               established. A box is provided in which to type the queries
               and the engine returns a result list which users can
               evaluate. The autocomplete suggestions assist users in
               defining their problems, however there is a lack of support
               for an iterative manual refinement of the query. This
               additional aid can be beneficial when users not know the
               exact terms to describe the concept they are looking for.
               Therefore, the goal of this project is to show searchers how
               a slight variation of the query changes the results. With
               this information, they then can perform a targeted
               refinement of the query to access useful information
               sources. To achieve this goal, each part of the searcher’s
               query is varied with a thesaurus that provides synonyms for
               the individual query terms. While performing the user’s
               original query in a normal fashion, variations of this query
               are conducted in the background. To provide a concise visual
               summary of the query results, text mining techniques are
               performed on all gathered results to retrieve the most
               important key terms for each query variation. This procedure
               results in a visual overview of what the searcher’s query
               finds together with what could be found with a slight
               variation of the query. Additional queries should make users
               aware that alternative queries may be more appropriate when
               their original query is poorly formulated. In conjunction
               with some interaction tools, the goal is to reduce the
               burden of refining search queries and therefore making
               searching the web less complex.",
  URL =        "https://www.cg.tuwien.ac.at/research/publications/2017/mazurek-2017-vows/",
}