System, method, and computer program for operating a multi-stage artificial intelligence (AI) analysis in a communication network

A method, comprising:

• acquiring, using a processor of a communication network, first log data of network activity of the communication network based on a first set of monitoring rules, wherein the first log data is of a lower level of resolution by including a first number of log events associated with a particular length of time;
• detecting a network situation in log data by an artificial intelligence (AI) system using an AI-model, wherein the log data includes at least the first log data;
• computing a confidence level associated with the detection of the network situation, wherein the confidence level corresponds to the AI-model used to detect the network situation;
• determining that the confidence level does not surpass a predefined value;
• responsive to determining that the confidence level does not surpass the predefined value, acquiring second log data of the network activity of the communication network based on an additional set of monitoring rules, wherein the second log data is of higher level of resolution by including a second number of log events associated with the particular length of time where the second number is greater than the first number;
• and
• repeating the detection of the network situation, the computation of the confidence level, and the determination whether the confidence level surpasses the predefined value, based on the second log data.

A device, comprising:

• a non-transitory memory storing instructions;
• and
• one or more processors in communication with the non-transitory memory, wherein the one or more processors execute the instructions to:
• acquire first log data of network activity of a communication network based on a first set of monitoring rules, wherein the first log data is of a lower level of resolution by including a first number of log events associated with a particular length of time;
• detect a network situation in log data by an artificial intelligence (AI) system using an AI-model, wherein the log data includes at least the first log data;
• compute a confidence level associated with the detection of the network situation wherein the confidence level corresponds to the AI-model used to detect the network situation;
• determine that the confidence level does not surpass a predefined value;
• responsive to determining that the confidence level does not surpass the predefined value, acquire second log data of the network activity of the communication network based on an additional set of monitoring rules, wherein the second log data is of higher level of resolution by including a second number of log events associated with the particular length of time where the second number is greater than the first number;
• and
• repeat the detection of the network situation, the computation of the confidence level, and the determination whether the confidence level surpasses the predefined value, based on the second log data.