A. Quamar, Fatma Özcan, Konstantinos Xirogiannopoulos
{"title":"知识库丰富实体的发现与创建","authors":"A. Quamar, Fatma Özcan, Konstantinos Xirogiannopoulos","doi":"10.1145/3214708.3214712","DOIUrl":null,"url":null,"abstract":"Businesses and professional organizations from a variety of different domains such as finance, weather, healthcare, social networks, etc., produce massive amounts of unstructured, semi-structured and structured data. Knowledge bases, enable querying and analysis of integrated content derived from such data available as open, third party and propriety data sets. Many knowledge bases today, provide an entity-centric view over the integrated content by using domain-specific ontologies. These entity-centric views enable querying individual real-world entities, as well as exploring exact information (such as address or net revenue of a company) through explicit querying using languages such as SQL or SPARQL. Although very useful for many business and commercial applications, this may not be sufficient for the exploration of relevant and context specific information associated with real-world entities stored in these knowledge bases. Users often need to resort to a manual and tedious process of exploration using ad-hoc queries to gather the required information. To enhance user experience and ameliorate the problem of relevant data exploration, we propose the concept of Rich Entities. These rich entities comprise of all the relevant and context specific information grouped together around real-world entities and served as efficient and meaningful responses to user queries against these entities in a knowledge base. These rich entities are created by grouping together information not only from a single entity represented as an ontology concept, but also related concepts and properties as specified by the domain ontology. In this paper we propose several novel techniques and algorithms to automatically detect, learn, and create domain-specific rich entities. We use inputs from query patterns in existing query workloads against knowledge bases, and leverage the structure and relationships between entities defined in the domain ontology. Our techniques are very effective and can be applied to a wide variety of application domains thus adding great value to data exploration and information extraction from entity-centric real-world knowledge bases.","PeriodicalId":93360,"journal":{"name":"Proceedings of the 5th International Workshop on Exploratory Search in Databases and the Web. International Workshop on Exploratory Search in Databases and the Web (5th : 2018 : Houston, Tex.)","volume":null,"pages":null},"PeriodicalIF":0.0000,"publicationDate":"2018-06-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"3","resultStr":"{\"title\":\"Discovery and Creation of Rich Entities for Knowledge Bases\",\"authors\":\"A. Quamar, Fatma Özcan, Konstantinos Xirogiannopoulos\",\"doi\":\"10.1145/3214708.3214712\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Businesses and professional organizations from a variety of different domains such as finance, weather, healthcare, social networks, etc., produce massive amounts of unstructured, semi-structured and structured data. Knowledge bases, enable querying and analysis of integrated content derived from such data available as open, third party and propriety data sets. Many knowledge bases today, provide an entity-centric view over the integrated content by using domain-specific ontologies. These entity-centric views enable querying individual real-world entities, as well as exploring exact information (such as address or net revenue of a company) through explicit querying using languages such as SQL or SPARQL. Although very useful for many business and commercial applications, this may not be sufficient for the exploration of relevant and context specific information associated with real-world entities stored in these knowledge bases. Users often need to resort to a manual and tedious process of exploration using ad-hoc queries to gather the required information. To enhance user experience and ameliorate the problem of relevant data exploration, we propose the concept of Rich Entities. These rich entities comprise of all the relevant and context specific information grouped together around real-world entities and served as efficient and meaningful responses to user queries against these entities in a knowledge base. These rich entities are created by grouping together information not only from a single entity represented as an ontology concept, but also related concepts and properties as specified by the domain ontology. In this paper we propose several novel techniques and algorithms to automatically detect, learn, and create domain-specific rich entities. We use inputs from query patterns in existing query workloads against knowledge bases, and leverage the structure and relationships between entities defined in the domain ontology. Our techniques are very effective and can be applied to a wide variety of application domains thus adding great value to data exploration and information extraction from entity-centric real-world knowledge bases.\",\"PeriodicalId\":93360,\"journal\":{\"name\":\"Proceedings of the 5th International Workshop on Exploratory Search in Databases and the Web. International Workshop on Exploratory Search in Databases and the Web (5th : 2018 : Houston, Tex.)\",\"volume\":null,\"pages\":null},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2018-06-15\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"3\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Proceedings of the 5th International Workshop on Exploratory Search in Databases and the Web. International Workshop on Exploratory Search in Databases and the Web (5th : 2018 : Houston, Tex.)\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1145/3214708.3214712\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Proceedings of the 5th International Workshop on Exploratory Search in Databases and the Web. International Workshop on Exploratory Search in Databases and the Web (5th : 2018 : Houston, Tex.)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1145/3214708.3214712","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
Discovery and Creation of Rich Entities for Knowledge Bases
Businesses and professional organizations from a variety of different domains such as finance, weather, healthcare, social networks, etc., produce massive amounts of unstructured, semi-structured and structured data. Knowledge bases, enable querying and analysis of integrated content derived from such data available as open, third party and propriety data sets. Many knowledge bases today, provide an entity-centric view over the integrated content by using domain-specific ontologies. These entity-centric views enable querying individual real-world entities, as well as exploring exact information (such as address or net revenue of a company) through explicit querying using languages such as SQL or SPARQL. Although very useful for many business and commercial applications, this may not be sufficient for the exploration of relevant and context specific information associated with real-world entities stored in these knowledge bases. Users often need to resort to a manual and tedious process of exploration using ad-hoc queries to gather the required information. To enhance user experience and ameliorate the problem of relevant data exploration, we propose the concept of Rich Entities. These rich entities comprise of all the relevant and context specific information grouped together around real-world entities and served as efficient and meaningful responses to user queries against these entities in a knowledge base. These rich entities are created by grouping together information not only from a single entity represented as an ontology concept, but also related concepts and properties as specified by the domain ontology. In this paper we propose several novel techniques and algorithms to automatically detect, learn, and create domain-specific rich entities. We use inputs from query patterns in existing query workloads against knowledge bases, and leverage the structure and relationships between entities defined in the domain ontology. Our techniques are very effective and can be applied to a wide variety of application domains thus adding great value to data exploration and information extraction from entity-centric real-world knowledge bases.