Data partioning based on end user behavior
End user data partitioning can include receiving a number of data queries for a data source from a user, developing a dimension relation graph based on attributes of the number of data queries, and partitioning the data source based on the dimension relation graph.
1. A method for end user behavior, data partitioning, comprising:
receiving a number of data queries for a data source from a user;
developing a dimension relation graph based on attributes of the number of data queries by generating a number of nodes corresponding to the attributes and a number of vertices linking the nodes, the vertices representing data queries that include attributes of the nodes linked by the respective vertices;
weighting each vertex, among the number of vertices, based on a quantity of data queries including attributes of the nodes linked by the respective vertex;
determining, from the dimension relation graph, a portion of the attributes that are most frequently utilized together among the number of data queries based on weights of the vertices of the dimension relation graph; and
partitioning the data source based on the dimension relation graph weights of the vertices so that data with the portion of the attributes is located in a same partition of the data source.
2. The method of claim 1 , wherein the number of nodes each represent a particular number of attributes.
3. The method of claim 2 , wherein generating the number of nodes includes weighting vertices between each node based on a user frequency.
4. The method of claim 2 , wherein generating the number of nodes includes linking the number of nodes based on the number of data queries.
5. The method of claim 2 , wherein weighting each node is based on a significance value within the data source.
6. The method of claim 1 , wherein partitioning the data source includes dynamically partitioning the data source utilizing changes in the dimension relation graph.
7. A non-transitory machine-readable medium storing a set of instructions executable by a processor to cause a computer to:
receive a number of data queries for a data source from a user that result in a particular report to the user;
develop a dimension relation graph by:
generating a number of nodes comprising attributes of the number of data queries,
weighting each node, among the number of nodes, based on a quantity of data within the data source that corresponds to the attributes,
linking the number of nodes, utilizing a number of vertices, based on the number of data queries, the vertices representing data queries that include attributes of the nodes linked by the respective vertices, and
weighting the number of vertices based on numbers of queries that include same attributes of the nodes linked by the respective vertices;
determine, from the dimension relation graph, a portion of the attributes that are most frequently utilized together among the number of data queries based on the weighted number of vertices; and
partition the data source based on the dimension relation graph so that data with the portion of the attributes is located in a same partition of the data source.
8. The medium of claim 7 , wherein each of the number of nodes are given a weight based on a frequency of the attributes within the data source.
9. The medium of claim 7 , wherein the instructions are executable to partition the data source includes instructions executable to partition the data source in a number of balanced partitions based on the dimension relation graph.
10. The medium of claim 7 , wherein the number of vertices represent a relation between attributes within the linked number of nodes.
11. The medium of claim 10 , wherein the relation occurs when attributes within each of the linked number of nodes is within the same query.
12. A system for end user data partitioning, the system comprising a processing resource in communication with a non-transitory machine readable medium, wherein the non-transitory machine readable medium includes a set of instructions and wherein the processing resource is designed to carry out the set of instructions to:
receive a number of data queries for a data source from a user, wherein each of the number of data queries includes a number of attributes;
generate a number of nodes based on the number of attributes, wherein each of the number of nodes relates to data from a different partition within the data source;
weight each node, among the number of nodes, based on a quantity of data within the data source that corresponds to the number of attributes;
link the number of nodes utilizing a number of vertices based on the number of data queries to develop a dimension relation graph, wherein the number of vertices are given a weight based on a frequency of the number of data queries that comprise the linked number of nodes;
determine, from the dimension relation graph, a portion of the attributes that are most frequently queried together among the number of data queries based on the weight of the number of vertices; and
partition the data source based on the dimension relation graph so that data with the portion of the attributes is located in a same partition of the data source.
13. The comp ing system of claim 12 , including instructions to dynamically give the number of vertices an updated weight based on real time frequency of the number of data queries.
14. The computing system of claim 12 , including instructions to give each of the number of nodes a weight based on a frequency of attributes within the data source.
15. The computing system of claim 12 , wherein each of the number of nodes comprises attributes from a different partition of the data source.