This document describes the MISP taxonomy format, a simple JSON format used to represent machine tags (also known as triple tags) vocabularies. A public directory, known as MISP taxonomies, is available and utilizes the MISP taxonomy format. These taxonomies are employed to classify cybersecurity events, threats, suspicious events, or indicators.
Sharing threat information has become a fundamental requirement in the Internet security and intelligence community at large. This information can include indicators of compromise, malicious file indicators, financial fraud indicators, or even detailed information about a threat actor. Classification plays a crucial role while sharing such indicators or information, ensuring adequate distribution, understanding, validation, or action regarding the shared information. The MISP taxonomies are a public repository of known vocabularies that can be utilized in threat information sharing.
In the MISP taxonomy context, machine tags help analysts to classify their cybersecurity events, indicators or threats. MISP taxonomies can be used for classification, filtering, triggering actions or visualisation depending on their use in threat intelligence platforms such as MISP [@?MISP-P].
Machine tags are represented as a string. Below listed are a set of sample machine tags for different namespaces such as tlp, admiralty-scale and osint.
The MISP taxonomy format uses the JSON [@!RFC8259] format. Each namespace is represented as a JSON object with meta information including the following fields: namespace, description, version, type.
namespace defines the overall namespace of the machine tag. The namespace is represented as a string and **MUST** be present. The description is represented as a string and **MUST** be present. A version is represented as a unsigned integer **MUST** be present. A type defines where a specific taxonomy is applicable and a type can be applicable at event, user or org level. The type is represented as an array containing one or more type and **SHOULD** be present. If a type is not mentioned, by default, the taxonomy is applicable at event level only. An exclusive boolean property **MAY** be present and defines at namespace level if the predicates are mutually exclusive.
predicates defines all the predicates available in the namespace defined. predicates is represented as an array of JSON objects. predicates **MUST** be present and **MUST** at least content one element.
The predicates array contains one or more JSON objects which lists all the possible predicates. The JSON object contains two fields: value and expanded. value **MUST** be present. expanded **SHOULD** be present. value is represented as a string and describes the predicate value. The predicate value **MUST** not contain spaces or colons. expanded is represented as a string and describes the human-readable version of the predicate value. An exclusive property **MAY** be present and defines at namespace level if the values are mutually exclusive.
The values array contain one or more JSON objects which lists all the possible values of a predicate. The JSON object contains two fields: predicate and entry. predicate is represented as a string and describes the predicate value. entry is an array with one or more JSON objects. The JSON object contains two fields: value and expanded. value **MUST** be present. expanded **SHOULD** be present. value is represented as a string and describes the machine parsable value. expanded is represented as a string and describes the human-readable version of the value.
colour fields **MAY** be used at predicates or values level to set a specify colour that **MAY** be used by the implementation. The colour field is described as an RGB colour fill in hexadecimal representation.
description fields **MAY** be used at predicates or values level to add a descriptive and human-readable information about the specific predicate or value. The field is represented as a string. Implementations **MAY** use the description field to improve more contextual information. The description at the namespace level is a **MUST** as described above.
: The Detection Maturity Level (DML) model is a capability maturity model for referencing ones maturity in detecting cyber attacks. It's designed for organizations who perform intel-driven detection and response and who put an emphasis on having a mature detection program.
: Gray Zone of Active defense includes all elements which lay between reactive defense elements and offensive operations. It does fill the gray spot between them. Taxo may be used for active defense planning or modeling.
: The Admiralty Scale or Ranking (also called the NATO System) is used to rank the reliability of a source and the credibility of an information. Reference based on FM 2-22.3 (FM 34-52) HUMAN INTELLIGENCE COLLECTOR OPERATIONS and NATO documents.
: The AIS Marking Schema implementation is maintained by the National Cybersecurity and Communication Integration Center (NCCIC) of the U.S. Department of Homeland Security (DHS)
: A series of assessment predicates describing the analyst capabilities to perform analysis. These assessment can be assigned by the analyst him/herself or by another party evaluating the analyst.
approved-category-of-action:
: A pre-approved category of action for indicators being shared with partners (MIMIC).
: Collaborative intelligence support language is a common language to support analysts to perform their analysis to get crowdsourced support when using threat intelligence sharing platform like MISP. The objective of this language is to advance collaborative analysis and to share earlier than later.
: The COPINE Scale is a rating system created in Ireland and used in the United Kingdom to categorise the severity of images of child sex abuse. The scale was developed by staff at the COPINE (Combating Paedophile Information Networks in Europe) project. The COPINE Project was founded in 1997, and is based in the Department of Applied Psychology, University College Cork, Ireland.
course-of-action:
: A Course Of Action analysis considers six potential courses of action for the development of a cyber security capability.
crowdsec:
: Crowdsec IP address classifications and behaviors taxonomy.
: It is critical that the CSIRT provide consistent and timely response to the customer, and that sensitive information is handled appropriately. This document provides the guidelines needed for CSIRT Incident Managers (IM) to classify the case category, criticality level, and sensitivity level for each CSIRT case. This information will be entered into the Incident Tracking System (ITS) when a case is created. Consistent case classification is required for the CSIRT to provide accurate reporting to management on a regular basis. In addition, the classifications will provide CSIRT IM’s with proper case handling procedures and will form the basis of SLA’s between the CSIRT and other Company departments.
: Cyber Threat Framework was developed by the US Government to enable consistent characterization and categorization of cyber threat events, and to identify trends or changes in the activities of cyber adversaries. https://www.dni.gov/index.php/cyber-threat-framework
: Taxonomy used by CyCAT, the Universal Cybersecurity Resource Catalogue, to categorize the namespaces it supports and uses.
cytomic-orion:
: Taxonomy to describe desired actions for Cytomic Orion
dark-web:
: Criminal motivation and content detection the dark web: A categorisation model for law enforcement. ref: Janis Dalins, Campbell Wilson, Mark Carman. Taxonomy updated by MISP Project and extended by the JRC (Joint Research Centre) of the European Commission.
: Taxonomy defined in the DCSO MISP Event Guide. It provides guidance for the creation and consumption of MISP events in a way that minimises the extra effort for the sending party, while enhancing the usefulness for receiving parties.
: Distributed Denial of Service - or short: DDoS - taxonomy supports the description of Denial of Service attacks and especially the types they belong too.
: German (DE) Government classification markings (VS).
death-possibilities:
: Taxonomy of Death Possibilities
deception:
: Deception is an important component of information operations, valuable for both offense and defense.
dga:
: A taxonomy to describe domain-generation algorithms often called DGA. Ref: A Comprehensive Measurement Study of Domain Generating Malware Daniel Plohmann and others.
: The Diamond Model for Intrusion Analysis establishes the basic atomic element of any intrusion activity, the event, composed of four core features: adversary, infrastructure, capability, and victim.
diamond-model-for-influence-operations:
: The diamond model for influence operations analysis is a framework that leads analysts and researchers toward a comprehensive understanding of a malign influence campaign by addressing the socio-political, technical, and psychological aspects of the campaign. The diamond model for influence operations analysis consists of 5 components: 4 corners and a core element. The 4 corners are divided into 2 axes: influencer and audience on the socio-political axis, capabilities and infrastructure on the technical axis. Narrative makes up the core of the diamond.
: A subset of Information Security Marking Metadata ISM as required by Executive Order (EO) 13526. As described by DNI.gov as Data Encoding Specifications for Information Security Marking Metadata in Controlled Vocabulary Enumeration Values for ISM
: Economical impact is a taxonomy to describe the financial impact as positive or negative gain to the tagged information (e.g. data exfiltration loss, a positive gain for an adversary).
: The present threat taxonomy is an initial version that has been developed on the basis of available ENISA material. This material has been used as an ENISA-internal structuring aid for information collection and threat consolidation purposes. It emerged in the time period 2012-2015.
: Estimative language to describe quality and credibility of underlying sources, data, and methodologies based Intelligence Community Directive 203 (ICD 203) and JP 2-0, Joint Intelligence
: EU classified information (EUCI) means any information or material designated by a EU security classification, the unauthorised disclosure of which could cause varying degrees of prejudice to the interests of the European Union or of one or more of the Member States.
: Reasons why an event has been extended. This taxonomy must be used on the extended event. The competitive analysis aspect is from Psychology of Intelligence Analysis by Richard J. Heuer, Jr. ref:http://www.foo.be/docs/intelligence/PsychofIntelNew.pdf
failure-mode-in-machine-learning:
: The purpose of this taxonomy is to jointly tabulate both the of these failure modes in a single place. Intentional failures wherein the failure is caused by an active adversary attempting to subvert the system to attain her goals – either to misclassify the result, infer private training data, or to steal the underlying algorithm. Unintentional failures wherein the failure is because an ML system produces a formally correct but completely unsafe outcome.
: Flesch Reading Ease is a revised system for determining the comprehension difficulty of written material. The scoring of the flesh score can have a maximum of 121.22 and there is no limit on how low a score can be (negative score are valid).
: The Future of Privacy Forum (FPF) [visual guide to practical de-identification](https://fpf.org/2016/04/25/a-visual-guide-to-practical-data-de-identification/) taxonomy is used to evaluate the degree of identifiability of personal data and the types of pseudonymous data, de-identified data and anonymous data. The work of FPF is licensed under a creative commons attribution 4.0 international license.
: Taxonomy related to the REGULATION (EU) 2016/679 OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL on the protection of natural persons with regard to the processing of personal data and on the free movement of such data, and repealing Directive 95/46/EC (General Data Protection Regulation)
: Information needed to track or monitor moments, periods or events that occur over time. This type of information is focused on occurrences that must be tracked for business reasons or represent a specific point in the evolution of ‘The Business’.
gea-nz-entities:
: Information relating to instances of entities or things.
gea-nz-motivators:
: Information relating to authority or governance.
: Updated (CIRCL, Seamus Dowling and EURECOM) from Christian Seifert, Ian Welch, Peter Komisarczuk, ‘Taxonomy of Honeypots’, Technical Report CS-TR-06/12, VICTORIA UNIVERSITY OF WELLINGTON, School of Mathematical and Computing Sciences, June 2006, http://www.mcs.vuw.ac.nz/comp/Publications/archive/CS-TR-06/CS-TR-06-12.pdf
ics:
: FIRST.ORG CTI SIG - MISP Proposal for ICS/OT Threat Attribution (IOC) Project
: How an incident is classified in its process to be resolved. The taxonomy is inspired from NASA Incident Response and Management Handbook. https://www.nasa.gov/pdf/589502main_ITS-HBK-2810.09-02%20%5bNASA%20Information%20Security%20Incident%20Management%5d.pdf#page=9
: A taxonomy describing information leaks and especially information classified as being potentially leaked. The taxonomy is based on the work by CIRCL on the AIL framework. The taxonomy aim is to be used at large to improve classification of leaked information.
information-origin:
: Taxonomy for tagging information by its origin: human-generated or AI-generated.
: Classification based on different categories. Based on https://www.sans.org/reading-room/whitepapers/incident/malware-101-viruses-32848
misinformation-website-label:
: classification for the identification of type of misinformation among websites. Source:False, Misleading, Clickbait-y, and/or Satirical News Sources by Melissa Zimdars 2019
: Malware Type and Platform classification based on Microsoft's implementation of the Computer Antivirus Research Organization (CARO) Naming Scheme and Malware Terminology. Based on https://www.microsoft.com/en-us/security/portal/mmpc/shared/malwarenaming.aspx, https://www.microsoft.com/security/portal/mmpc/shared/glossary.aspx, https://www.microsoft.com/security/portal/mmpc/shared/objectivecriteria.aspx, and http://www.caro.org/definitions/index.html. Malware families are extracted from Microsoft SIRs since 2008 based on https://www.microsoft.com/security/sir/archive/default.aspx and https://www.microsoft.com/en-us/security/portal/threat/threats.aspx. Note that SIRs do NOT include all Microsoft malware families.
: Malware Type and Platform classification based on Microsoft's implementation of the Computer Antivirus Research Organization (CARO) Naming Scheme and Malware Terminology. Based on https://www.microsoft.com/en-us/security/portal/mmpc/shared/malwarenaming.aspx, https://www.microsoft.com/security/portal/mmpc/shared/glossary.aspx, https://www.microsoft.com/security/portal/mmpc/shared/objectivecriteria.aspx, and http://www.caro.org/definitions/index.html. Malware families are extracted from Microsoft SIRs since 2008 based on https://www.microsoft.com/security/sir/archive/default.aspx and https://www.microsoft.com/en-us/security/portal/threat/threats.aspx. Note that SIRs do NOT include all Microsoft malware families.
mwdb:
: Malware Database (mwdb) Taxonomy - Tags used across the platform
: The taxonomy is meant for large scale cybersecurity incidents, as mentioned in the Commission Recommendation of 13 September 2017, also known as the blueprint. It has two core parts: The nature of the incident, i.e. the underlying cause, that triggered the incident, and the impact of the incident, i.e. the impact on services, in which sector(s) of economy and society.
nis2:
: The taxonomy is meant for large scale cybersecurity incidents, as mentioned in the Commission Recommendation of 13 May 2022, also known as the provisional agreement. It has two core parts: The nature of the incident, i.e. the underlying cause, that triggered the incident, and the impact of the incident, i.e. the impact on services, in which sector(s) of economy and society.
: Open Threat Taxonomy v1.1 base on James Tarala of SANS http://www.auditscripts.com/resources/open_threat_taxonomy_v1.1a.pdf, https://files.sans.org/summit/Threat_Hunting_Incident_Response_Summit_2016/PDFs/Using-Open-Tools-to-Convert-Threat-Intelligence-into-Practical-Defenses-James-Tarala-SANS-Institute.pdf, https://www.youtube.com/watch?v=5rdGOOFC_yE, and https://www.rsaconference.com/writable/presentations/file_upload/str-r04_using-an-open-source-threat-model-for-prioritized-defense-final.pdf
: After an incident is scored, it is assigned a priority level. The six levels listed below are aligned with NCCIC, DHS, and the CISS to help provide a common lexicon when discussing incidents. This priority assignment drives NCCIC urgency, pre-approved incident response offerings, reporting requirements, and recommendations for leadership escalation. Generally, incident priority distribution should follow a similar pattern to the graph below. Based on https://www.us-cert.gov/NCCIC-Cyber-Incident-Scoring-System.
: PyOTI automated enrichment schemes for point in time classification of indicators.
ransomware:
: Ransomware is used to define ransomware types and the elements that compose them.
ransomware-roles:
: The seven roles seen in most ransomware incidents.
retention:
: Add a retenion time to events to automatically remove the IDS-flag on ip-dst or ip-src attributes. We calculate the time elapsed based on the date of the event. Supported time units are: d(ays), w(eeks), m(onths), y(ears). The numerical_value is just for sorting in the web-interface and is not used for calculations.
: Runtime or software packer used to combine compressed or encrypted data with the decompression or decryption code. This code can add additional obfuscations mechanisms including polymorphic-packer or other obfuscation techniques. This taxonomy lists all the known or official packer used for legitimate use or for packing malicious binaries.
scrippsco2-fgc:
: Flags describing the sample
scrippsco2-fgi:
: Flags describing the sample for isotopic data (C14, O18)
: The Targeted Threat Index is a metric for assigning an overall threat ranking score to email messages that deliver malware to a victim’s computer. The TTI metric was first introduced at SecTor 2013 by Seth Hardy as part of the talk “RATastrophe: Monitoring a Malware Menagerie” along with Katie Kleemola and Greg Wiseman.
: Thales Group Taxonomy - was designed with the aim of enabling desired sharing and preventing unwanted sharing between Thales Group security communities.
threatmatch:
: The ThreatMatch Sectors, Incident types, Malware types and Alert types are applicable for any ThreatMatch instances and should be used for all CIISI and TIBER Projects.
threats-to-dns:
: An overview of some of the known attacks related to DNS as described by Torabi, S., Boukhtouta, A., Assi, C., & Debbabi, M. (2018) in Detecting Internet Abuse by Analyzing Passive DNS Traffic: A Survey of Implemented Systems. IEEE Communications Surveys & Tutorials, 1–1. doi:10.1109/comst.2018.2849614
: The Traffic Light Protocol (TLP) (v2.0) was created to facilitate greater sharing of potentially sensitive information and more effective collaboration. Information sharing happens from an information source, towards one or more recipients. TLP is a set of four standard labels (a fifth label is included in amber to limit the diffusion) used to indicate the sharing boundaries to be applied by the recipients. Only labels listed in this standard are considered valid by FIRST. This taxonomy includes additional labels for backward compatibility which are no more validated by FIRST SIG.
: The Indicator of Trust provides insight about data on what can be trusted and known as a good actor. Similar to a whitelist but on steroids, reusing features one would use with Indicators of Compromise, but to filter out what is known to be good.
