Font Size: a A A

Keywords-based Temporal Information Retrieval Method Over Relation Databases

Posted on:2018-01-16Degree:MasterType:Thesis
Country:ChinaCandidate:X M ZhangFull Text:PDF
GTID:2348330512477229Subject:Computer Science and Technology
Abstract/Summary:PDF Full Text Request
With the coming of the Large Data Era,a large number of accumulated data has become an important data assets of in all walks of life.As the time dimension becoming an important indicator which represents the value of data,the temporal data has been attracting more and more attention.So how to effectively store,manage and retrieve temporal data has become a focus in the field of database and information retrieval.Database information retrieval is a technology between database and information retrieval,which can effectively support ordinary users to query keywords on the database efficiently.Great achievements have been obtained in the technology of database information retrieval.The research in the field of temporal information retrieval shows that by incorporating the temporal information into the information retrieval technology,the temporal query can be efficiently processed and the information needed by the user can be quickly and efficiently retrieved.However,the keywords retrieval methods of relational database take the data temproality into no consideration,so the retrieval of temporal data is rare.To solve this problem,this paper studies the temporal retrieval method of keywords on relational databases from the time dimension.Firstly,the relevant theories of temporal information processing are introduced,which provide thought and theoretical support for the research of temporal retrieval method.Then,based on the original relational database keyword search method,the time dimension is introduced to propose the temporal information retrieval method based on keywords-based information retrieval method over relation databases.The method consists of three parts:firstly,the temporal data graph is constructed by analyzing the temporal relation between entities and temporal entities stored in the database.Secondly,as the existing indexing methods can not satisfy the search of the fast temporal keyword nodes,the temporal inverted index has been designed,which makes the temporal partition of the temporal node sets corresponding to each keyword,and improves the searching efficiency.Thirdly,a temporal retrieval algorithm T-STAR is designed.T-STAR algorithm mainly uses the time pruning strategy,which is to prun the edges that don't meet the constraints in the retrieval,so that the retrieval results can satisfy the time constraints of temporal queries.At the same time,the method of calculating the temporal edge weight is proposed,which can better meet the content relevance of the retrieval results.Finally,the prototype system of relational database temporal retrieval based on keywords has been implemented.The validity of the method has been validated by using the existing temporal datasets such as Employees and NBA.The experimental results have been evaluated with P@K and MAP.Experiment has shown that the method can effectively improve the effect of database information retrieval and satisfy the users' temporal retrieval requirements under the precondition of ensuring efficiency.
Keywords/Search Tags:Temporal Data Graph, Temporal Inverted Index, Temporal Information Retrieval, Relation Databases, Keyword Retrieval
PDF Full Text Request
Related items