摘要:
Embodiments of the present disclosure set forth a method for retrieving information from a semantic database having a plurality of semantic data. The method includes selecting a hub that is resolved with a number of semantic data based on a resolution rule, when the number of the semantic data is greater than a threshold in an offline environment; calculating a semantic data set that is associated with the hub based on a searching approach and the resolution rule in the offline environment; indexing the semantic data set in the offline environment; modifying the semantic database to include the semantic data set in the offline environment; and retrieving information from the semantic data set in an online environment.
摘要:
Modularization-based reasoning of reasoning tasks is provided. Performing a reasoning task includes receiving the reasoning task. Symbols in the reasoning task are interpreted to extract a module from the knowledge base. The reasoning task is them performed with the extracted module to obtain a result of the reasoning task.
摘要:
A database system and method allows a user to write program code in a high-level programming language that implements a class that defines the structure of a user-defined type and methods that can be invoked on instances of the type. The class is then registered with the database system, which enforces a specific contract for user-defined types against the class. The contract comprises the following requirements. First, the class must specify one of a plurality of different formats for persisting instances of the user-defined type in a database store. Second, the class must be capable of returning a null value for the user-defined type. Third, the class must provide a method for converting the user-defined type to another type. Once these requirements are satisfied, the database system enables instances of the user-defined type to be created. A user-defined type can be used in all contexts where any other built-in type can be used. This includes defining columns of tables, declaring variables and parameters to functions and stored procedures, and evaluating expressions of the user-defined type in SQL queries and updates. Preferably, the database system stores metadata about the user-defined type for subsequent use in creating instances of the type.
摘要:
Any string in any character set with an arbitrary-leveled weight-based comparison system is transformed into a bitstring in such a way that two transformed strings can be compared byte-by-byte. The resulting bit string has the minimum possible maximum length. The transformed bit strings can be inverted—meaning the original string can be recovered from the transformed string.
摘要:
Method and implementations for providing a secure data storage service in a cloud computing environment are generally disclosed. The method comprises: partitioning a data resource into data particles, assigning logic groups to the data particles, assigning physical storage groups to the data particles, and/or storing each physical storage group at corresponding storage resource, receiving a request for the data resource, determining whether the request for the data resource is valid, and if the request is valid, transmitting the data particles of the data resource to the client. The method enables improved security for accessing data, and also improves the user experience in cloud computing environments.
摘要:
Techniques for ontology expansion are provided. In one disclosed method, an ontology is first identified. Terms in the ontology are disambiguated to identify term senses. Additional senses are identified using a lexical database. The additional senses are then added to the ontology to generate an augmented or expanded ontology.
摘要:
A method of searching for objects in a second database using a query from a first database includes receiving a first query from the first database where the first query is generated using a collation associated with the first database or user session or explicitly specified by users, where the first database may be the same as the second database. The first query is rewritten to form a second query which includes a superset of the collation and a residue predicate which includes a predicate from the first query. An index plan is adopted which targets an index associated with the a multiple language collation expressed in the rewritten query. The rewritten query is executed accessing the index and corresponding data that align with the residue predicate. The total number of indexes provided can be much less than the number of indexes needed if all possible collations were individually indexed.