In data management, semantic warehousing is a methodology of digitalized text data using similar functions to Data warehousing (DW), such as ETL(Extract, transform, load), ODS(Operational data store), and MODEL. Key value operation is less useful for the digitalized text. Semantic warehousing is different from DW in that semantic information base from text(semantic) data.
Semantic warehousing is different from search engine in that semantic information base from text data is stored in the database.(DBMS)
Though data is most important word in computing era, it can not explain human knowledge well yet. Data(numeric data) is key element of computing systems for certain organization (especially companies, enterprises), but no performance oriented organization needs something to gather and use knowledge or human feeling. Semantic warehousing will be equally or more important than data warehousing in the future.
Semantic warehousing is a conceptual and functional term meaning to gather from a source, semantically defining and providing information from digitalized text type data.
Data warehousing (DW) is popular these days. Gathering data from systems that generate transactions, data warehouses become a base of information. Key of data warehouse is a model (called datamart) and that model is made up of dimensions(key) and measures(value). Users get information from the models by doing certain operations. Online analytical processing (OLAP) is most the important operation for the users to get information from the DW models. Handling dimensions with pivoting, drilling, slice & dice operations users get numeric values like sales amounts, growth rates, etc. Various areas of this world defined and appeared on the World Wide Web(Internet), eager to present their contents in a semantic way. Briefly speaking semantic warehousing has datawarehousing boby and search head and ontology features.
Data warehousing contributed to companies’ business values and lots of solutions and tools are commercially successful. Analysis of internal data delivers a certain level of business values, on the contrary to this Semantic warehousing environment has not yet matured. Capacity of social data is increasing rapidly and various efforts of finding value from that data are made widely known as Big data, etc. Semantic warehousing can be the mainstream of treat data and intelligence of social world in the future though it is defined with other keywords.
At the Big data era, semantic processing is going to become major IT process. Semantic warehousing is digital infra of Intelligence.
▣ Medical area (Clinical Information)
Some hospital implement semantic warehousing for clinical information (SWCI). Medical information is now knowledge network level. UMLS define semantic knowledge network of medical language. Currently medical information stored in database and not fully used for clinic. Semantic warehousing is next stage of digitalized medical information.
SWCI is a name of conceptual system of clinical information.
Named by Juhan Kim (SNUH, Seoul National University Hospital) and Bohyon Hwang, YongChan Keum in 2008.
Defined architecture on SWCI ;
1. Semantic-oriented cleansing
2. Semantic-oriented meta management
3. Clinical(Medical) knowledge basement
4. Semantic-oriented user intelligence
▣ Intelligence Area
At the point of Big data usage, intelligence reporting can be valuable results.
- Source information
2. Manage intelligence & Semantic data
3. Intelligence service & use
– Big data
– Semantic web
– Medical and healthcare : EMR (Electronic Medical Record), EHR (Electronic Health Record)
– Data warehouse
– AI (artificial intelligence)
- BILaboratory of Seoul National University Hospital
- Smith, Barry Kumar, Anand and Schulze-Kremer, Steffen (2004) Revising the UMLS Semantic Network, in M. Fieschi, et al. (eds.), Medinfo 2004, Amsterdam: IOS Press, 1700.