Abstract: Method, system, and programs for heterogeneous data management. Information from multiple data sources is first obtained. Data/metadata from each of the data sources is modeled based on the source and/or granularity information of the data/metadata to generate data/metadata models. The data/metadata from multiple data sources are integrated, by applying one or more processes to the data/metadata from different data sources based on the data/metadata models, to generate integrated data/metadata. A provenance representation for the integrated data/metadata is created tracing sources, granularities, and/or processes applied and archived for enabling an query associated with the integrated data/metadata.

Abstract: Methods and systems are provided that may be utilized for spilling in query processing environments.

Abstract: Method, system, and programs for heterogeneous data management. Information from multiple data sources is first obtained. Data/metadata from each of the data sources is modeled based on the source and/or granularity information of the data/metadata to generate data/metadata models. The data/metadata from multiple data sources are integrated, by applying one or more processes to the data/metadata from different data sources based on the data/metadata models, to generate integrated data/metadata. A provenance representation for the integrated data/metadata is created tracing sources, granularities, and/or processes applied and archived for enabling an query associated with the integrated data/metadata.

Abstract: Techniques described herein adaptively select materialized view fragments for persistent maintenance. During an interval of time, the selected fragments are persistently maintained in the database system, while the other non-selected fragments are not persistently maintained as materialized view fragments. Over time, the composition of the set of selected fragments may change. As queries are executed in the database system over an interval of time, statistics including the frequency of access of each currently selected fragment during that interval are generated. At the start of the next interval of time, based on these statistics, some currently selected fragments may be unselected. Some currently non-selected fragments of one or more candidate materialized views may be selected based on the statistics.

Abstract: A system of query scheduling to maximize work sharing. The system schedules queries to account for future queries possessing a sharability component. Included in the system are operations for assigning an incoming query to a query queue based on a sharability characteristic of the incoming query, and evaluating a priority function for each member of a plurality of query queues to identify one highest priority query queue. The priority function accounts for the probability that a future incoming query will contain the sharability characteristic common to a member of the plurality of query queues. The system of query scheduling to maximize work sharing selects a batch of queries from the highest priority query queue, and dispatches the batch to one or more query execution units.

Abstract: Techniques described herein adaptively select materialized view fragments for persistent maintenance. During an interval of time, the selected fragments are persistently maintained in the database system, while the other non-selected fragments are not persistently maintained as materialized view fragments. Over time, the composition of the set of selected fragments may change. As queries are executed in the database system over an interval of time, statistics including the frequency of access of each currently selected fragment during that interval are generated. At the start of the next interval of time, based on these statistics, some currently selected fragments may be unselected. Some currently non-selected fragments of one or more candidate materialized views may be selected based on the statistics.

Abstract: In a distributed system that includes multiple machines, a scheduler attempts to schedule a task on a machine that is not currently overloaded with work. If a task is scheduled on a machine that does not yet have copies of the portions of the data set on which the task needs to operate, then that machine obtains copies of those portions from other machines that already have them. Whenever a “source” machine ships a copy of a portion to another “destination” machine in the distributed system, the destination machine persistently stores that copy on the destination machine's persistent storage mechanism. The copy also remains on the source machine. Thus, portions of the data set are automatically replicated whenever those portions are shipped between machines of the distributed system. Each machine in the distributed system has access to “global” information that indicates which machines have which portions of the data set.

Abstract: A system of query scheduling to maximize work sharing. The system schedules queries to account for future queries possessing a sharability component. Included in the system are operations for assigning an incoming query to a query queue based on a sharability characteristic of the incoming query, and evaluating a priority function for each member of a plurality of query queues to identify one highest priority query queue. The priority function accounts for the probability that a future incoming query will contain the sharability characteristic common to a member of the plurality of query queues. The system of query scheduling to maximize work sharing selects a batch of queries from the highest priority query queue, and dispatches the batch to one or more query execution units.

Abstract: Techniques are provided for combining searching and browsing into a single interface. The topology and content of a document collection or web site is determined and user information need keywords are determined. Relevant documents are identified using search techniques and the relevant document paths determined. The relevant document paths are analyzed using information scent to generate information scent values for the relevant document paths. A display attribute of the associated connection or link is then synthesized based on the information scent, the keywords and the document. The process repeats for each new document or web page.

Abstract: Techniques are provided for combining searching and browsing into a single interface. The topology and content of a document collection or web site is determined and user information need keywords are determined. Relevant documents are identified using search techniques and the relevant document paths determined. The relevant document paths are analyzed using information scent to generate information scent values for the relevant document paths. A display attribute of the associated connection or link is then synthesized based on the information scent, the keywords and the document. The process repeats for each new document or web page.