Research On Granularity Modeling And Temporal Relations Of Uncertain Temporal Information

For these recent years, information systems in all walks of life have accumulated a large number of data thanks to the rapid development of information technology. Temporal data is one of the most important data, which exists in the application fields of news systems, medical information systems and transportation systems etc., and plays increasingly significant roles in many research fields, such as temporal data mining, temporal knowledge and temporal information processing. Uncertain temporal information is important content of temporal information, and it is gradually being known as a research focus in the temporal domain and artificial intelligence. The issue of uncertain temporal information is the core problem of semantic understanding, temporal planning, uncertain temporal logic, temporal data mining and spatial information processing, as well as in the fields of artificial intelligence and natural language processing.Uncertain temporal information includes the modification of prepositions, and time, interval or span with granularity attribute. It belongs to the imprecise concept without a priori knowledge. How to make the computer understand these uncertain temporal information based on semantics, process its uncertain semantic, make approximate calculation, and further construct the universal temporal granularity model and explore the qualitative and quantitative temporal relations? These problems are exactly the difficult points in the field of temporal information processing.This study proposed the conversion idea of uncertain semantic, converting it to a neighborhood or interval to be a calculated problem, aiming at the uncertainty and diversity caused by the uncertain semantic; put forward a formal description of the granule point, grain interval and span of uncertain temporal to make the uncertain temporal element be involved in calculation, combining the granularity attribute and uncertain semantics; with the idea of interval set and rough set, reconstructed the temporal elements in discrete state, and used the thought of lower approximation and upper approximation to clearly divide the uncertain and certain elements in uncertain temporal, according to the uncertain temporal structure; thereby put forward the calculation method of approximate accuracy of uncertain temporal granule point,interval and span, at the same time, relevant examples to descript it is given.Given the uncertain temporal modeling problem, this study put forward the tuple granularity model of uncertain time point, interval and span from uncertain temporal semantics, combining the constraint of temporal granularity; and made generalization processing on uncertain temporal to provide the premise for investigating the uncertain temporal relations.Given the problem of uncertain temporal relations, this study explored the uncertain temporal relations using both qualitative and quantitative method on the basis of the generalization processing, focused on the quantitative calculation of the temporal relations with complex overlapping:and multiple possibilities and proved its completeness, and represented and dealt with the uncertain temporal information in medical information systems in the form of examples. In specific, this paper proposed the operation among uncertain temporal spans and semantics reduction method combined temporal interval number, and thus effectively dealt with the complex operation among uncertain temporal spans including multiplication, addition and subtraction.The main innovation of this paper lied in:(1) proposed corresponding conversion and portrayal methods to make uncertain temporal involve in the operation, given the diverse uncertain semantic; (2) reconstructed uncertain temporal and proposed the approximate accuracy computation on the uncertain temporal; (3) constructed the tuple granularity model of uncertain time, interval and span and made generalization processing on uncertain temporal to provide the premise for investigating the uncertain temporal relations; (4) described the uncertain temporal relations with qualitative method, explored the temporal relations with complex overlapping using quantitative method and proved its completeness. Specifically put forward its approximate operation method of uncertain temporal spans and dealt with the complex operation among them.