DY IQX: A ovel Meta-Search Engine for the Web

The effect of metadata in collection fusion has not been sufficiently studied. In response to this, we present a novel meta-search engine called Dyniqx for metadata based search. Dyniqx integrates search results from search services of documents, images, and videos for generating a unified list of ranked search results. Dyniqx exploits the availability of metadata in search services such as PubMed, Google Scholar, Google Image Search, and Google Video Search etc for fusing search results from heterogeneous search engines. In addition, metadata from these search engines are used for generating dynamic query controls such as sliders and tick boxes etc which are used by users to filter search results. Our preliminary user evaluation shows that Dyniqx can help users complete information search tasks more efficiently and successfully than three well known search engines respectively. We also carried out one controlled user evaluation of the integration of six document/image/video based search engines (Google Scholar, PubMed, Intute, Google Image, Yahoo Image, and Google Video) in Dyniqx. We designed a questionnaire for evaluating different aspect of Dyniqx in assisting users complete search tasks. Each user used Dyniqx to perform a number of search tasks before completing the questionnaire. Our evaluation results confirm the effectiveness of the meta-search of Dyniqx in assisting user search tasks, and provide insights into better designs of the Dyniqx’ interface.