An Intelligent SIEM–SOAR Framework Using Isolation Forest

Security Operations Centres (SOCs) are flooded with huge numbers of events flowing from network and authentication sources, conventional monitoring based on rules fails to cope with unusual activity, and complete automation of response can lead to disruption of legitimate services. This paper proposes a Intelligent SIEM-SOAR framework which integrates unsupervised anomaly detection and human supervised incident response. Three data sources are ingested: CICIDS2017 network flows, Suricata EVE-format intrusion events and Linux SSH authentication logs. The CICIDS2017 records are pre-processed using the following steps: cleaned, reduced to 12 behavioural features, standardized and scored by an Isolation Forest trained on a benign baseline. The detections are then transformed into structured alerts with anomaly scores, severity ratings, managed using an Elasticsearch–Logstash–Kibana (ELK) configuration, and enriched with MITRE ATT&CK techniques, orchestrated via a Shuffle-compatible orchestration client. A React dashboard allows a SOC analyst to view alerts, mark a positive or negative outcome, and approve or disapprove proposed containment – actions approved are executed only in a simulated sandbox and added to an audit trail, which can be used later for retraining. Consideration for deployment of a candidate Isolation Forest model involves a promotion gate to test the candidate model against the baseline. The precision of the baseline model on the bundled test subset was 84.27 %, recall 100 %, F1 score 91.46 % and accuracy 94.40 % on a bundled test set of 500 test and 500 baseline records. The reported candidate model evaluation gave an F1-score of 94.04 % and achieved a decrease in false positives from 28 to 19 while preserving the recall rate. But the archive corresponding to the same evaluation artefact recorded no analyst feedback samples used, so it cannot be said that the improvement was due to analyst feedback. IndexTerms - SIEM, SOAR, Isolation Forest, Anomaly detection, CICIDS2017, Security Operations Centre, Suricata, MITRE ATT&CK.

Authors

Institutions

Publication Details

Journal
International Journal of Creative and Open Research in Engineering and Management
Published
2026-11-13
DOI
https://doi.org/10.55041/ijcope.v2i11.049
Primary Topic
Network Security and Intrusion Detection
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

An Intelligent SIEM–SOAR Framework Using Isolation Forest

D. Tejaswini, Deevi Radha Rani, B. Naveen Kumar, Ch. Giri Manish Kumar et al.
2 links
International Journal of Creative and Open Research in Engineering and Management
Network Security and Intrusion Detection
article

An Intelligent SIEM–SOAR Framework Using Isolation Forest

D. Tejaswini, Deevi Radha Rani, B. Naveen Kumar, Ch. Giri Manish Kumar, B. Varshitha
article en

Abstract

Security Operations Centres (SOCs) are flooded with huge numbers of events flowing from network and authentication sources, conventional monitoring based on rules fails to cope with unusual activity, and complete automation of response can lead to disruption of legitimate services. This paper proposes a Intelligent SIEM-SOAR framework which integrates unsupervised anomaly detection and human supervised incident response. Three data sources are ingested: CICIDS2017 network flows, Suricata EVE-format intrusion events and Linux SSH authentication logs. The CICIDS2017 records are pre-processed using the following steps: cleaned, reduced to 12 behavioural features, standardized and scored by an Isolation Forest trained on a benign baseline. The detections are then transformed into structured alerts with anomaly scores, severity ratings, managed using an Elasticsearch–Logstash–Kibana (ELK) configuration, and enriched with MITRE ATT&CK techniques, orchestrated via a Shuffle-compatible orchestration client. A React dashboard allows a SOC analyst to view alerts, mark a positive or negative outcome, and approve or disapprove proposed containment – actions approved are executed only in a simulated sandbox and added to an audit trail, which can be used later for retraining. Consideration for deployment of a candidate Isolation Forest model involves a promotion gate to test the candidate model against the baseline. The precision of the baseline model on the bundled test subset was 84.27 %, recall 100 %, F1 score 91.46 % and accuracy 94.40 % on a bundled test set of 500 test and 500 baseline records. The reported candidate model evaluation gave an F1-score of 94.04 % and achieved a decrease in false positives from 28 to 19 while preserving the recall rate. But the archive corresponding to the same evaluation artefact recorded no analyst feedback samples used, so it cannot be said that the improvement was due to analyst feedback. IndexTerms - SIEM, SOAR, Isolation Forest, Anomaly detection, CICIDS2017, Security Operations Centre, Suricata, MITRE ATT&CK.

International Journal of Creative and Open Research in Engineering and ManagementVol. 02(11)
GMR Institute of Technology (IN)
Network Security and Intrusion Detection
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon