Generating a semantic graph relating information assets using feedback re-enforced search and navigation
Embodiments of the invention provide an approach for creating, evolving and using a weighted semantic graph to manage and potentially identify certain information assets within an enterprise. The semantic graph may be generated by monitoring users navigating through search results which provide a set of information assets responsive to a search query. By recording the navigation path taken by many users, relationships between information assets may be identified. Further, once generated, the semantic graph may be used to present users with in indication of related information assets as part of the search results. Further still, the semantic graph may also be used to identify information assert “hubs” as well as information assets that may provide low utility to individuals within the enterprise.
1. A non-transitory computer-readable storage medium storing an application, which, when executed on a processor, performs an operation for identifying relationships between a plurality of information assets managed by an enterprise, the operation comprising:
monitoring a plurality of users navigating paths through respective search results provided by the application;
generating, from the monitoring, a semantic graph capturing a set of relationships between information assets, of the plurality of information assets, based on a sequence of information assets selected by the plurality of users when accessing the respective search results; and
wherein monitoring the plurality of users and generating the semantic graph includes:
assigning a weight to a relationship between respective first and second information assets navigated by the plurality of users in the respective search results, wherein the weight assigned to a given first and second information asset is increased each time one of the plurality of users navigates from the given first information asset to the second information asset identified in the respective search results, and
upon determining the assigned weight exceeds a specified threshold, adding an edge to the semantic graph to connect nodes in the semantic graph corresponding to the given first and second information assets.
2. The non-transitory computer-readable storage medium of claim 1 , wherein each respective search result identifies a subset of the plurality of information assets managed by the enterprise.
3. The non-transitory computer-readable storage medium of claim 1 , wherein the semantic graph includes a plurality of nodes, each corresponding to one of the plurality of information assets and includes a plurality of weighted edges, each indicating a relationship strength between two information assets connected by a respective edge.
4. The non-transitory computer-readable storage medium of claim 1 , wherein the operation further comprises:
identifying at least a first node in the semantic graph having zero edges to other nodes; and
prompting an administrator to assign one or more edges to the first node.
5. The non-transitory computer-readable storage medium of claim 1 , wherein the information assets include at least one of a database system, a web service and a server application.
6. A system, comprising:
a processor; and
a memory storing an enterprise information asset management application program, which, when executed on the processor, performs an operation for identifying relationships between a plurality of information assets managed by an enterprise, the operation comprising:
monitoring a plurality of users navigating paths through respective search results provided by the enterprise information asset management application,
generating, from the monitoring, a semantic graph capturing a set of relationships between information assets, of the plurality of information assets, based on a sequence of information assets selected by each of the plurality of users accessing the respective search results, and
wherein monitoring the plurality of users and generating the semantic graph includes:
assigning a weight to a relationship between respective first and second information assets navigated by the plurality of users in the respective search results, wherein the weight assigned to a given first and second information asset is increased each time one of the plurality of users navigates from the given first information asset to the second information asset identified in the respective search results; and
upon determining the assigned weight exceeds a specified threshold, adding an edge to the semantic graph to connect nodes in the semantic graph corresponding to the given first and second information assets.
7. The system of claim 6 , wherein each respective search result identifies a subset of the plurality of information assets managed by the enterprise.
8. The system of claim 6 , wherein the semantic graph includes a plurality of nodes, each corresponding to one of the plurality of information assets and includes a plurality of weighted edges, each indicating a relationship strength between two information assets connected by a respective edge.
9. The system of claim 6 , wherein the operation further comprises:
identifying at least a first node in the semantic graph having zero edges to other nodes; and
prompting an administrator to assign one or more edges to the first node.
10. The system of claim 6 , wherein the information assets include at least one of a database system, a web service and a server application.