Method, system and program product for allocation and/or prioritization of electronic resources

A method comprising:

• determining or obtaining or receiving, by one or more computers, a first rule set comprising a plurality of patterns of run-time data used to distinguish among a plurality of applications and/or locations within applications running on monitored computers associated with respective user IDs and to provide respective application context IDs that designate respective applications incorporating application locations within applications;
• receiving or obtaining or being presented, by the one or more computers, with run-time data for a first plurality of the user IDs, wherein the runtime data for a respective one of the user IDs comprises and/or is derived from run-time instantiations of objects occurring on the monitored computer for the respective user ID, with the objects comprising one or more selected from the group of methods, system objects, object events, dynamic GUI objects, network component objects, wherein the object events comprise at least one selected from the group of operating system events, application events, and web browser events and network events generated or affecting directly or indirectly the respective monitored computer;
• determining, by the one or more computers, respective application context IDs for applications and/or respective locations within applications running on the respective user IDs of a second plurality of the user IDs, based at least in part on the first rule set;
• detecting, by the one or more computers, a pattern comprising one or more selected from the group of an application, a combination of applications, and a combination of an application used by a specified one of the user IDs;
• determining a prediction rule for the determined pattern, by the one or more computers, from a prediction rule set, wherein the respective prediction rule in the prediction rule set is based at least in part on a past history of usage of the electronic resources when the pattern is detected;
• determining, by the one or more computers, prediction data for of electronic resources that may be used at a future time or during a window of time when the pattern is detected based at least in part on the prediction rule set;
• and
• using the prediction data for electronic resources, by the one or more computers, in one selected from the group of generating of the allocation data for the electronic resources and providing data or making the data accessible electronically that indicates a need for additional electronic resources.

A method, comprising:

• receiving or obtaining, by the one or more computers, slowdown data indicating a slow down of a first one of the monitored computers;
• determining, by the one or more computers, a resource shift rule applicable at a time of receiving or obtaining the slowdown data, the resource shift rule for taking electronic resources from one or more other monitored computers, the resource shift rule comprising a shift pattern comprising one or more selected from the group of a time of day, a day of the month, a company department, an application, wherein the respective resource shift rule in the resource shift rule set is based at least in part on a past history of usage of the electronic resources when the shift pattern is detected;
• determining, by the one or more computers, using the resource shift rule the one or more other of the monitored computers from which to take the electronic resources;
• and
• generating automatically allocation data, by the one or more computers, for shifting additional of the electronic resources to the first monitored computer, to be taken at least in part from the one or more other monitored computers only on a temporary basis and then returning the electronic resources allocated to the first monitored computer and the other monitored computers to previous levels before the shifting.

A method comprising:

• creating or receiving, by one or more computers, a language syntax to define at least in part:
• a plurality of rules for assigning token types to a stream of messages to obtain a stream of contextualized token instances;
• a plurality of production rules for detecting deterministic event instances from the stream of contextualized token instances;
• a plurality of state type rules for detecting state type instances, wherein a state type rule comprises one or more deterministic event types with respective probability weights, wherein a probability weight is the likelihood that, detecting the deterministic event type that results in the system, the respective state type is executed;
• and
• a plurality of respective pattern rules to define respective Markov chains, with each of the pattern rules comprising a set of state types and transition probability weights between the state types;
• receiving or obtaining or being presented, by the one or more computers, with the run-time data comprising a stream of messages for systems of a first plurality of the user IDs, wherein the runtime data for a respective one of the user IDs comprises and/or is derived from run-time instantiations of objects occurring on the monitored computer for the respective user ID, with the objects comprising one or more selected from the group of methods, system objects, object events, dynamic GUI objects, network component objects, wherein the object events comprise at least one selected from the group of operating system events, application events, and web browser events and network events generated or affecting directly or indirectly the respective monitored computer;
• parsing, by the one or more computers, to apply the language syntax to the stream of messages to extract a stream of the contextualized token instances from the stream of messages based at least in part on the language syntax;
• detecting, by the one or more computers, from the stream of the token instances, a stream of the deterministic user event instances, based at least in part on the language syntax;
• selecting, by the one or more computers, a plurality of the Markov chains containing one or more of the state type instances that are implied from the stream of messages, based at least in part on the respective probability weights of the state types from the deterministic user event instances that were detected;
• computing, by the one or more computers, a probability that the observed system is executing the respective pattern type for each of the respective Markov chains selected;
• and
• determining, by the one or more computers, based at least in part on one or more of the selected Markov chains and next state transitions in the respective Markov chains selected and/or one or more of predicted states in the respective Markov chains selected, predicted prioritization data for electronic resources for the observed system and/or a predicted level of electronic resources that may be needed by the system generating the messages.