中文摘要：

就目前物联网中的传感器查询方法而言，存在不能有效地利用语义信息和对用户掌握的查询技术的高度依赖性两个主要问题。为了解决以上问题，提出一种本体和签名图结合的传感器查询方法。根据传感器本体的知识结构，将关键字查询转换成更为准确的结构化查询，这样用户就不需要了解具体的转换细节和掌握相关的查询技术；构建签名图后创建查询树，用来缩小查询范围。最后通过与其他经典的传感器查询方法进行比较，结果表明提出的方法在传感器查询准确率和效率方面有一定的优势。

英文摘要：

In terms of current sensor query methods in Internet of things, there are two major problems, which are the disability to effectively use of semantic information and the high-dependency on detailed query skills for users. To solve above-mentioned problems, this paper developed a sensor query method combined ontology and signature graph. Firstly, the method transformed a keyword query to a more accurate structured query according to sensor ontology, it reduced the heavy burden that users knew the specific details of transformation and relevant query technology. Secondly, it built a query tree after constructing signature graphs to significantly reduce the query scope based on this query tree. Finally, the comparison with other classical sensor query methods reveals that the proposed method has some special advantages in respects of accuracy and efficiency.