Visual Information Retrieval Interfaces

研究人员对信息检索(IR)的多个参考点模型的研究导致了一系列视觉信息检索界面(VIRI)的发展，这些界面体现了可视化文档之间相似性和语义关系的不同方法。这些VIRI旨在帮助用户检索和组织在线信息。基于布尔/集合运算的可视化(BIRD)、由参考点定义的简化空间中的距离测量(GUIDO)、相对于多个参考点的相似性(VIBE)、以及由距离和角度间隔(DARE)共同定义的相似性，已经开发出VIRI。将对这些VIRI进行用户研究，以便：1)改进其界面的可用性，2)调查不同可视化对潜在用户群体的概念性可访问性，以及3)评估VIRI对典型信息检索和组织任务的效用。这些研究将首次对文件空间可视化对各种信息检索和组织任务的预期用户群体的有用性进行系统调查。该界面的特点是增强了Viri的可用性，信息可视化展示了其有用性，预计将增强超文本作为组织和访问在线信息的机制。

The investigators' research in multiple reference point models of information petrieval (IR) has led to the development of a collection of visual information retrieval interfaces (VIRIs) embodying different approaches to visualizing similarity and semantic relations among documents. These VIRIs are intended to aid users in retrieval and organization of online information. VIRIs based on visualization of Boolean/set operations (BIRD), distance measures in a reduced space defined by reference points (GUIDO), similarity with respect to multiple reference points (VIBE), and similarity defined jointly by distance and angular separation (DARE) have been developed. User studies will be conducted with these VIRIs in order to: 1) improve usability of their interfaces, 2) investigate conceptual accessibility of different visualizations to potential user populations, and 3) evaluate utility of the VIRIs for prototypical information retrieval and organization tasks. These studies will provide the first systematic investigation of the usefulness of document space visualizations to expected user populations for a variety of information retrieval and organization tasks. The interface features enhancing VIRI usability, and information visualizations demonstrating usefulness are expected to augment hypertext as mechanisms for organizing and accessing online information.