International Business Machines Corporation

Healtech Software India Pvt. Ltd.

Kyndryl, Inc.

Atri Mandal

Vijay Ekambaram

Gargi Banerjee Dasgupta

Giriprasad Sridhara

Sudhir Shetty

Pooja Aggarwal

Nikhil Malhotra

Raja Shekhar Mulpuri

Palavali Shravan Kumar Reddy

Ruchi Mahindru

Ajay Gupta

Anbang Xu

Anindya Neogi

Amitkumar Manoharrao Paradkar

Sriram Raghavan

Daniela Rosu

Atul Kumar

Xiaotong Liu

Prateeti Mohapatra

Animesh Nandi

Suriya Subramanian

Shubham Atreja

Howard Zhang

Pujitha Kara

Shainy Merin

Adityam Ghosh

Jaisri S

Arpit Rathi

An aspect of the present disclosure facilitates measuring the capability of AIOps (Artificial Intelligence for IT operations) systems deployed in computing environments. In one embodiment, a first simulation of a target AIOps system is run using a first historical input set having a corresponding first actual output set of a first AIOps system different from the target AIOps system. A second simulation of a reference AIOps system is run using a second historical input set having a corresponding second actual output set of the same first AIOps system. A first and second accuracy scores are determined based on outputs of the first and second simulations and the corresponding first and second actual output sets. An enablement score representing a measure of the capability (in terms of accuracy of prediction) of the target AIOps system is generated based on the first accuracy score and the second accuracy score.

Measuring the capability of AIOps systems deployed in computing environments

Proactive avoidance of performance issues in computing environments. In one embodiment, a causal dependency graph representing the usage dependencies among the various components of a computing environment is formed, the components being associated with key performance indicators (KPIs). A probabilistic model is trained with prior incidents that have occurred in the components to correlate outliers of KPIs in associated components to prior incidents. The training includes determining the correlation based on the causal dependency graph. Upon detecting the occurrence of outliers for performance metrics, an imminent performance issue likely to occur in a specific component is identified based on the probabilistic model and the detected outliers. A preventive action is performed to avoid the occurrence of the imminent performance issue in the specific component.

Proactive avoidance of performance issues in computing environments using a probabilistic model and causal graphs

An aspect of the present disclosure provides a relevance ranking system for events identified by performance management systems. In one embodiment, a system receives, from one or more performance management systems, events associated with performance metrics and determines, based on the received events, a corresponding relevance score for each performance metric. The system generates a ranking list of the received events based on the corresponding relevance scores (determined for the performance metrics). The system then provides the ranking list of events to a user (such as an administrator). According to another aspect, the system (noted above) performs the steps of receiving, determining, and generating for each time interval of a set of time intervals.

Relevance ranking system for events identified by performance management systems

According to an aspect, a (recommendation) system constructs a knowledge graph based on problem descriptors and remediation actions contained in multiple incident reports previously received from a performance management (PM) system. Each problem descriptor and remediation action in an incident report are represented as corresponding start node and end node in the knowledge graph, with a set of qualifier entities in the incident report represented as causal links between the start node and the end node. Upon receiving an incident report related to an incident identified by the PM system, the system extracts a problem descriptor and a set of qualifier entities. The system traverses the knowledge graph starting from a start node corresponding to the extracted problem descriptor using the set of qualifier entities to determine end nodes representing a set of remediation actions. The system provides the set of remediation actions as recommendations for resolving the incident.

Recommending remediation actions for incidents identified by performance management systems

Embodiments are provided that relate to a computer system, a computer program product, and a computer-implemented method for improving performance of a virtual dialog agent system employing an automated virtual dialog agent. Embodiments involve generating ground truth (GT) from a user's knowledge base, and leveraging the GT to evaluate how the virtual dialog agent performs with the GT. The evaluation measures quality of a multi-turn virtual dialog, and generates a remediation plan directed at an algorithmic improvement of the virtual dialog agent.

Virtual dialog system performance assessment and enrichment

A cognitive assignment engine (CAE) system attempts to infer semantic meaning from textual content of an incoming message in order to use the inferred meaning to assign the message to an appropriate responder. If the message contains insufficient textual content, the system identifies ontological structures comprised by the message's graphical content and classifies each structure as a function of the structure's location within the graphical content or of an intrinsic characteristic of the structure. The system then generates a message identifier by performing a computation on these classifications and uses the identifier to retrieve a previously stored graphical template that comprises ontological structures similar to those of the incoming message. The system associates the incoming message with a semantic meaning previously associated with the template, enabling the system to classify the message and to assign the message to the correct responder.

Self-learning ontology-based cognitive assignment engine

A help desk management system uses segment partitioning and matching of each identified segment to a suitable problem and resolver group and then sequencing the partitioned segments based on sentiment analysis and sequence mining on historical tickets and audit logs to actuate effective resolution and pipelining of helpdesk tickets needing resolutions from multiple resolver groups. The helpdesk tickets can be in the form of e-mails.

Resolution and pipelining of helpdesk tickets needing resolutions from multiple groups

Methods, systems, and computer program products for fault localization for cloud-native applications are provided herein. A computer-implemented method includes classifying an event-related alert directed to at least one system by processing one or more characteristics of the event-related alert; obtaining and processing multiple application logs based at least in part on the classification of the event-related alert; identifying error logs among the multiple application logs based at least in part on the processing of the multiple application logs; ordering the error logs using one or more prioritization techniques; and performing at least one automated action based at least in part on the ordering of the error logs.

Fault localization for cloud-native applications

Aspects of the disclosure provide for a method. In at least some examples, the method includes receiving logs of a computer system, wherein a log comprises information regarding an operation of the computer system. The method also includes performing a course analysis of the logs to identify invariants, parameters, and gray area terms included in the logs, wherein gray area terms are terms included in the logs that are undefined between identifications of invariant or parameter by the course analysis. The method also includes performing a fine analysis to identify each of the gray area terms as either an invariant or a parameter to generate first log parsing templates, wherein the fine analysis is a sequence labeling analysis. The method also includes performing a similarity analysis to combine similar templates from among the first log parsing templates to form second log templates.

Log parsing template generation

One embodiment provides a method, including: receiving a foreign language trouble ticket requiring resolution; translating text of the foreign language trouble ticket into a language known to the person, wherein the translating comprises (i) translating a subset of foreign language keywords within a portion of the foreign language trouble ticket identified as a problem portion into the known language and (ii) translating a remaining subset of keywords into the known language using keyword links generated from previously resolved tickets by: extracting keywords from the historical tickets, wherein the keywords are recognized as corresponding to an identified portion; and generating at least one keyword link from at least one of the identified portions identified as a problem description portion; and directing the known language ticket to a resolver group, wherein the resolver group is selected based upon an issue identified within the ticket.

Translation of ticket for resolution

One embodiment provides a method for use within an organization having a plurality of groups dedicated to providing assistance with requests, including: obtaining a communication comprising a request for assistance from a user; identifying, using information contained within the obtained communication, a group that is able to provide assistance in resolving the request, each group having a corresponding model identifying information needed by that group; determining, based upon a model corresponding to the identified group, information needed by the identified group to resolve the request; identifying at least one modification that needs to be made to the communication based upon the determined information, wherein the identifying at least one modification comprises determining a difference between the determined information and the information contained within the obtained communication; and sending a modified communication to the identified group, wherein the modified communication comprises the obtained communication having at least one incorporated identified modification.

Communication enrichment recommendation

One embodiment provides a method, including: receiving a plurality of loglines from a system execution log; generating at least one control flow graph by mining (i) a plurality of templates and (ii) a sequence of the plurality of templates defined by an order encountered during execution from the system execution log, wherein the control flow graph identifies flows between templates during execution; mining (i) discriminative parameters, (ii) a position for each of the mined discriminative parameters within a template, and (iii) a value of each of the mined discriminative parameters; annotating edges of the at least one control flow graph with the values of the mined discriminative parameters within the control flow graph; identifying execution traces from the system execution log using the values of the mined discriminative parameters; and detecting system anomalies in the identified execution traces by comparing one of the execution traces to an expected execution.

System anomaly detection using parameter flows

Methods and arrangements for efficiently mining a control flow graph from execution logs of a distributed system. Using at least one text clustering technique, two text clusters are generated from the plurality of execution logs. At least one approximate template is generated based on the at least two text clusters. At least one refined template is created via refining the at least one approximate template using multimodal sequencing. The control flow graph is created based on the at least one refined template. An anomaly is detected in the control flow graph.

Growing community of inventors

Bengaluru, India

Atri Mandal