Time and Location: 4/24/2020, Friday, 4 p.m., Online seminar

Speaker: Mr. Chenxu Niu

Title: SeMIQS: Semantic Metadata Indexing and Querying Service for Self-describing File Formats

Abstract: It is a daunting and highly challenging task for researchers to find datasets relevant to their needs, especially for self-describing files, such as HDF5 and NetCDF, which is often used for storage in scientific applications. Existing solutions extract the metadata and process the querying with the aid of database management. However, they are hindered by capturing the semantic meanings of the content of the metadata, and therefore limited in performing queries at semantic level. In our research, we propose semantic metadata indexing and querying service(SeMIQS) for self-describing scientific datasets, which utilizes a word-vector model to capture semantics of the metadata in self-describing files and achieve metadata querying service at semantic level. By introducing metadata semantization approach and semantic metadata indexing mechanism, SeMIQS provides the capability to semantically match a group of keywords with a collection of self-describing files.