Accurate Context-Aware Search for Managed Document Collections

搜索是大量计算应用程序的关键组成部分。然而，虽然对通用文本的大型集合的搜索很容易理解，并且最好的系统非常有效，但对较小或特殊目的集合的搜索则不太可靠，并且吸引了相对较少的研究。通过确定使用上下文（如用户历史记录或先前页面有用性）的一般方法，可以大大提高这种情况下的搜索质量。利用这些原则的产品将提供更高的工作效率，并能够定位其他搜索工具无法识别的信息。

Search is a key component of a vast range of computing applications. However, while search on large collections of general-purpose text is well-understood, and the best systems are highly effective, search on smaller or special-purpose collections is much less reliable and has attracted relatively little research. By identifying general ways of using context such as user history or prior page usefulness, the quality of search in such cases can be greatly improved. Products that make use of these principles will provide greater workplace efficiency and be able to locate information that other search tools cannot identify.