Your browser does not support JavaScript. Please to enable it.

Terms & Conditions

The idea you wish to view belongs to a community that requires acceptance of terms and conditions.


    Help to Improve This Idea.



    Winning solution

    Preventing the Financing of Terrorism by Engineering an ML Model with Financial Crime Data

    By Rudradeb Mitra

    Challenge: "engineering a model for financial crime data correlation"

    goFintel is the “NEXT Generation Platform for detecting, preventing and countering money laundering and the financing of terrorism”

    In close collaboration with United Nations Secretariat substantive units and Member States User Communities (UC), the United Nations Office of Information and Communications Technology (OICT) has been delivering sustainable and affordable IT solutions and making them available to government institutions for use in their jurisdictions.

    goFintel (“government office Financial Intelligence”) is the next generation software platform being developed within the framework of the UN Office of Counter Terrorism “UNOCT-UNCCT Global coordinated programme on detecting, preventing and countering the financing of terrorism” addressing UN Security Council Resolution 2462 and 2482 mandates for Member States to strengthen their capacity to combat money laundering and the financing of terrorism.

    goFintel aims to enhance Member States capacity to collect, use, retain and transfer tactical and strategic intelligence by employing emerging technologies (Artificial Intelligence, Machine Learning and Blockchain Analysis etc.), while remaining in compliance with international and regional standards (e.g., FATF, EU AML Directives, Interpol, Europol etc).

    goFintel seeks to shift away from a siloed approach to intelligence and expand money laundering and terrorism financing investigation capabilities through analysis that spans multiple and different data sources. This approach enables Financial Intelligence Units (FIU) to rely on advanced information sharing, data transmission and retention, secure ICT architectures and cross-functional operational capabilities with multiple stakeholders (e.g., law enforcement, judiciary, etc.).


    Financial Crime Data Challenge Goal(s)
    Work on a model(s) to correlate data from financial, criminal, and other open data records to identify, prevent and counter the financing of terrorism. The expected result would be a comprehensive package describing the mechanisms, technologies, algorithms that can be used to correlate seemingly disparate open data sets to identify connections and patterns of financial crime activities.

    Whether you are an academic faculty student, an entrepreneur or an established company, or simply a creative person, you can help us build goFintel. Either teams or individual (self organized online and offline) Data Scientists, Data Engineers, AI Experts, Machine learning Engineers and Researchers are particularly encouraged to participate in the Challenge. In case of team work, a Project leader should be identified to coordinate the activities and to report to the goFintel Team.

    - Describe the model to correlate open data from various sectors using supervised and unsupervised learning systems (Machine/Deep Learning, Artificial Intelligence).
    - Identify options to display/visualize data in many formats
    - Establish a model to visualize data on demand and/or in real-time.

    Financial Crime Data
    - Open Source financial crime records/company registers, social media contents, OpenSource intelligence etc. available on the public web
    - United Nations Security Council Consolidated Lists, available here:

    An XML schema widely used by Reporting Entities to transmit Suspicious Transaction Report/Suspicious Activity Report (STR/SAR) to Financial Intelligence Units which describes data points matching reporting requirements is available here:

    Open Data Structure Typology
    - Structured, i.e. in a form of an existing database.
    - Semi (or Quasi) Structured, i.e. in a form of a structured or delimited text.
    - Unstructured, i.e. in form of a text document, pictures, sound, videos, machine logs, etc.

    Open Data Integration Options
    - On demand.
    - On batch-basis.
    - Real time.

    Open Datasets Characteristics
    1. Large Volume – related pictures, videos, and/or sound from social media could potentially have extremely large volumes.
    2. High Velocity – the real-time data would come at immense velocity requiring real-time or near-real-time processing capacity.
    3. High Variety – data comes in different structure formats (structured, semi-structured, and unstructured).
    4. Unknown Veracity – data has various levels of veracity; some veracity can be determined while other could be unknown.

    Expected Outcomes
    Participants will deliver one or more design proposal(s) to implement a software solutions on a relevant Open Source licence (GPL, MIT, Creative Commons etc) performing supervised and unsupervised learning (Machine/Deep Learning, Artificial Intelligence) using open source data.
    In addition, participants are required to share all the assets associated with their submission with the organizers (e.g. including code, data, and collateral materials) on a publicly-available online repository visible to the general public without requiring any authentication or password protection.

    Open Source
    All inputs and outputs of this project must be covered by recognized open source and creative commons licensing. OICT encourages the leveraging and extension of existing Open Source frameworks in compliance with their relevant intellectual property rights licensing provisions. In this regard, participants will be required to accept the terms and conditions approved by the UN Office of Legal Affairs regarding intellectual property, available at:

    Phases & Timeline
    The call for solutions will be open as of 1st April 2021. Throughout all phases described below, OICT encourages constant interaction with the goFintel Team to discuss the requirements for this assignment. Questions regarding this challenge can be posted in the comment section of Questions & Answers.

    i. Phase one: (deadline 30 July 2021) submission of a design proposal for an implementable solution relying on an identified subset of open data sources, along with related comprehensive approach references. Support the result with visual mock-ups and/or initial Proof of Concept and/or presentation of proposed solution design specifications.

    ii. Review phase, (deadline 31 August 2021). A panel composed of OICT, UNOCT UNCCT, National Financial Intelligence Units personnel and other partners will review and evaluate the proposed models.

    Review of Solutions & Evaluation Criteria
    The models received will be evaluated by a panel of OICT, UNOCT-UNCCT Officers and Representatives from a number of National Financial Intelligence Units with the objective to rank and identify the most relevant and well-developed tools. Proposed solutions will be reviewed based on the below criteria:

    • Relevance and applicability to the task objectives described in the above-mentioned principles.
    • Maturity and completeness.
    • Ease of use and user-friendliness.
    • Modularity, which allows integration with other products.
    • Maintainability and quality of the algorithms.
    • Documentation.

    Suggested participant hands-on expertise
    • Data collection, ingestion and storing of structured, semi-structured and un-structured data in real time, in a pre-determined batch process, or on demand.
    • Knowledge in Cross-Industry Process for Data Mining (CRISP-DM) methodology.
    • Data modelling advanced techniques, correlation, deep learning, analysis, reporting and alerting.
    • Software development (Python and R).
    • AI, Machine learning frameworks.

    The successful solution will be showcased by the submitter at the relevant goFintel Seminar. The target audience will include terrorism financing and financial intelligence experts, key national FIU personnel along with UN Officials involved in the programme.

    Unite Ideas – the “United Nations platform for innovating social impact”
    Unite Ideas is a global community where individuals, teams and organizations can collaborate with the United Nations to leverage innovative ideas and technology for sustainable development.

    Unite Ideas facilitates collaboration between the United Nations, academia, and civil society to advance the Sustainable Development Goals (SDGs), Human Rights, Peace and Security, and the Rule of Law through crowdsourcing challenges.

    The platform engages and mobilizes students, data scientists, programmers, designers and entrepreneurs worldwide to develop open source technology solutions to challenges posed by partners that are united in their goal to generate social good.
    Unite Ideas members can connect with United Nations staff and other global experts who are uniquely committed to positively impacting the world.

    {"currentPhase":"Complete","enddateepoch":1630443600,"days":-23,"completed":true,"phases":[{"phase":1,"name":"Pre-Launch testing","description":"Pre-Launch testing","progress":100,"percent":19},{"phase":2,"name":"Submit solution","description":"Submit your solution","progress":100,"percent":35},{"phase":3,"name":"Review","description":"Review of solutions","progress":100,"percent":8},{"phase":4,"name":"Submit solution","description":"Submit your solution","progress":100,"percent":22},{"phase":5,"name":"Winner Announcement","description":"Winner Announcement","progress":100,"percent":16}]}
    00 DAYS 00 HRS 00 MINS
    Challenge ended

    OICT logo

    OCT Logo

    goFintel logo

    No Ideas have been selected.
    • 6
    • 5
    • 1
    • 72