Crowdsourcing Implementation Based On Automatic Mapping Of Taxonomy

As an important tool for knowledge organization,taxonomy plays an important role in the classification of library bibliographic resources and document resources,as well as in the navigation and retrieval of various network information resources.At present,different countries and institutions use different taxonomy to index their respective information resources,and the corresponding relationships are not clearly established between different classification systems,thus affecting the efficiency of mutual visits.With the inter-library loan between libraries and the promotion of resource sharing,the exchange of visits between library bibliographic databases has become a trend.Establishing the mapping between different vocabularies is an important method to realize the interoperability of classification systems,and it is of great significance to the mapping between different taxonomy.The academic circles at home and abroad have actively explored and implemented the mapping of taxonomy.The mapping between taxonomy is divided into manual direct mapping and computer automatic mapping.The results of the manual mapping are accurate and complete,but the workload is large and the completion period is long.The automatic mapping of the computer improves the mapping efficiency,and the results are inaccurate.Some categories cannot establish mappings,and full automatic mapping cannot be realized.Combining manual mapping and automatic mapping can effectively improve the accuracy and efficiency of taxonomy mapping.Crowdsourcing is an innovative model to solve tasks that are difficult for machines to accomplish alone by integrating users and machine.The practice of introducing crowdsourcing into the mapping methods is to use the knowledge background and wisdom of the public to complete the mapping work,which can improve the accuracy of mapping and find more mapping categories.Based on the above background,this paper proposes an implementation method based on automatic mapping for crowdsourcing mapping.Construct the corresponding automatic mapping algorithm by analyzing the characteristics of the classification scheme.Based on the results of automatic mapping,the model of crowdsourcing mapping is designed to achieve the matching between different classification schemes.This paper first gives a brief overview of the basic theory and the mapping methods of taxonomy mapping.Select CLC?DDC?LCC which are widely used in the world,and construct a mapping between CLC and DDC,CLC and LCC.Analyzing the characteristics of the three taxonomies,the method based on category similarity is used to realize the mapping between CLC and DDC.The morphemes segmentation is performed in the mapping process.In the mapping between CLC and LCC,the mapping between DDC and LCC is introduced.The DDC and LCC mapping based on the associated bibliographic data is proposed.The DDC class table is used as an intermediary to indirectly establish the mapping relationship between CLC and LCC.Then,based on the automatically mapped data,the crowdsourcing taxonomy mapping mode is constructed by analyzing the user crowdsourcing mapping framework,the crowdsourcing mapping task module,and the crowdsourcing mapping quality control scheme.Finally,the accuracy and effectiveness of the method are verified by experiments and provide reference for the mapping of other taxonomies.