[目的/意义]信息检索处理的是相关性的不确定性问题，但在技术层面则通常将不确定性转化为确定性的处理方法，对信息内容中存在的不确定性语义关注不多，而这一问题在某些信息检索应用场景中可能显著地影响信息检索的结果，因此针对这类不确定性语义，需要考虑针对性的处理方法。[方法/过程]提出基于D-S证据理论的不确定性语义表示方法和将这类不确定性语义特征与文本特征、主题特征相融合的检索模型，并利用公开的数据集开展实验研究，对所提出的模型进行实验。[结果/结论]D-S理论中的证据区间概念能够描述上述不确定性，多源证据融合方法也能够将这类不确定性语义特征与文本特征、主题特征融合，并通过模型训练得出理想参数，进而改进检索结果。这一模型在理论上具有包容性与可扩展性，基于该模型融合其他检索方法是进一步需研究的内容。
[Purpose/significance] The uncertainty issue of relevance is important in IR, and is turned into certainty in different IR models. However, semantic uncertainty is common in the search query and documents, and this kind of uncertainty can significantly affect information retrieval in some situations, which needs special methods to process.[Method/process] An information model based on Dempster-Shafer Theory was proposed to represent semantic uncertainty and multi-source evidences including texts, topics and semantic uncertainty. An experiment was conducted to test that model.[Result/conclusion] It is found that evidence interval of Dempster-Shafer Theory can be used to represent semantic uncertainty, and the combining method of multi-evidence is capable to deal with texts, topics and semantic uncertainty.Ideal parameter can be generated from the training to improve retrieval results. This model is inclusive and extensive in theories, and using this model to combine other IR methods needs further research.
[1] 韩毅. 信息检索过程的不确定性分析[J]. 图书情报工作, 2005, 49(5):102-104.
[2] HELTON, J. C. Uncertainty and sensitivity analysis in the presence of stochastic and subjective uncertainty[J]. Journal of statistical computation and simulation, 1997,57:3-76.
[3] SMETS P, KENNES R. The transferable belief model[J]. Artificial intelligence, 1994, 66(2):191-234.
[4] YAGER R R, LIU L. Classic works of the Dempster-Shafer theory of belief functions[M]. Berlin:Springer, 2008:1-34.
[5] 王知津, 卞丹, 王文爽. 信息检索中"相关性"的概念、类型及其不确定性[J]. 山东图书馆学刊, 2011(1):1-5.
[6] BAEZA-YATES R A, RIBEIRO-NETO B. Modern information retrieval[M]. 北京:机械工业出版社, 2011.
[7] MANNING C D, RAGHAVAN P, SCHUTZE H.信息检索导论[M]. 王斌,译. 北京:人民邮电出版社, 2010.
[8] 段新生. 证据理论与决策、人工智能[M]. 北京:中国人民大学出版社, 1993.
[9] 潘巍, 王阳生, 杨宏戟. D-S证据理论决策规则分析[J]. 计算机工程与应用, 2004, 40(14):14-17.
[10] SHAFER G A. Mathematical theory of evidence[M]. Princeton:Princeton University Press,1976.
[11] YAGER R R. On the Dempster-Shafer framework and new combination rules[J]. Information sciences, 1987, 41(2):93-137.
[12] YAGER R R. Decision making under Dempster-Shafer uncertainties[M]//YAGER R R, LIU L.Classic works of the Dempster-Shafer theory of belief functions. Berlin:Springer, 2008:619-632
[13] LALMAS M. Information retrieval and Dempster-Shafer's theory of evidence[M]//HUNTER A, PARSONS S.Applications of uncertainty formalisms. Berlin:Springer, 1998:157-176.
[14] ZHANG J L, DENG B X, LI X. Using the Dempster-Shafer theory of evidence to rank documents[J]. 清华大学学报自然科学版(英文版), 2012, 17(3):241-247.
[15] DENG B X, Zhang J L, SHI C. A new document retrieval model using Dempster-Shafer theory of evidence[C]//International conference oninformation science and engineering(2009). Washington, DC:IEEE Computer Society, 2009:746-749.
[16] LALMAS M, RUTHVEN I. Representing and retrieving structured documents using the Dempster-Shafer theory of evidence:modelling and evaluation[J]. Journal of documentation, 1998, 54(5):529-565.
[17] OROUMCHIAN F, ARAABI B N, ASHOORI E. An application of plausible reasoning and Dempster-Shafer theory of evidence in information retrieval[C]//Proceedings of the international conference on fuzzy systems and knowledge discovery:computational intelligence for the e-age. Singapore:[s. n.],2002:442-447
[18] SIMPSON M S,VOORHEES E M, HERSH W.Overview of the TREC 2014 clinical decision support track[EB/OL].[2017-06-14]. http://trec.nist.gov/pubs/trec23/papers/overview-clinical.pdf.
[19] ROBERTS K,SIMPSON M,VOORHEES E, et al. Overview of the TREC 2015 clinical decision support track[EB/OL].[2017-06-14].http://trec.nist.gov/pubs/trec24/papers/Overview-CL.pdf.
