Skip to content

The-inevitables/Analysis-hub

 
 

Repository files navigation

Analysis Hub

Analysis Hub is a web-app, that has been designed to simplify data visualization easier so that it's more user relatable and customizable. We propose an app built on top of Elastic cloud and Kibana frameworks to analyze and provide a graphical representation of data related to traffik analysis in Malaysia, profiles for money laundering, and KYC due diligence profiling

The app will have a unique user account so that security is ensured for every user and the data they use. On logging in the user will be taken to the homepage where options to select any three of the outcomes mentioned, namely:

  • Traffik analysis - A dashboard dedicated to providing users with the unique and filed specific representation of each data field from the CSV file. The dashboard has visualizations for max values of victim gender across multiple locations in Malaysia, trafficking modes, and gender to trafficker ratio.

  • Palm Oil Analysis- This dashboard has been designed to represent the palm oil industry-specific data visualization ranging from trafficking from a recruitment perspective to data related to the companies' internal and external reports. It also represents data on a map with the palm oil industrial areas highlighted.

  • Profiles for money laundering. - A dashboard designed and built on a dataset that shows the type of frauds involved in the transaction, the type of action involved, and the dates of actual cases reported in hotspots with a higher rate of repetition in particular locations.

  • Due diligence and KYC - KYC dashboard has been built on top of sample dataset which represents incomplete profiles that are missing either PAN or Aadhar details or having incomplete profiles with missing names as such.

Once selecting one of the above options, the user will be given an option to input the datasets based on which the outcome results will be generated with the generic algorithm we use to predict the result.

Options for users to download the result and also the graphical representation of the obtained results will be displayed within the application. Users will have the option to view and modify the previous results or analysis performed by the user on the previous data page.

Once the results are obtained users will be provided with the “get legal help” option, which will let the user learn about the laws and regulations related to each of the sectors in which they are trying to obtain the analysis report. Also, they can take the issue to any legal organization with the downloadable reports generated by the application.

Traffic Analysis identifier

This section generally deals with producing a traffic analysis data virtualization based on the input provided. The main features are :

  • The user will be able to upload the necessary data as single or multiple files.
  • The system will automatically model the user profile from the given set of data.
  • A graphical and statistical report will be shown as the end result of the modeling
  • AccountUser can download the result and share it across platforms for further requirements.
  • The account user will have the provision to manually filter out data from the previous modeling done on the system.
  • External encryption can be done on the downloadable format form of the report to increase the security of the data.
  • A dashboard dedicated to providing users with the unique and filed specific representation of each data field from the CSV file. The dashboard has visualizations for max values of victim gender across multiple locations in Malaysia, trafficking modes, and gender to trafficker ratio.

Sector and geographic-specific risk profiling

This section generally deals with producing a specific risk profile based on the inputs given, normally based on the geographical factors provided. The main features are:

  • The user will be able to upload the necessary data as single or multiple files
  • The system will automatically model the user profile from the given set of data.
  • A graphical and statistical report will be shown as the end result of the modeling
  • AccountUser can download the result and share it across platforms for further requirements.
  • The account user will have the provision to manually filter out data from the previous modeling done on the system.
  • External encryption can be done on the downloadable format form of the report to increase the security of the data

Profiles for money laundering

This section generally deals with flagging and finding out money laundering accounts based on the inputs given, normally based on the geographical factors provided. The main features are:

  • The AccountUser can upload the input files on the web app with the necessary parameters that need to be modeled on the system. This can either be a CSV or other large data files.
  • The system will create a model based on the input given and produce an output that will highlight the money laundering accounts and their specific category of money laundering
  • The AccountUser will get a graphical and statistical representation of the output
  • There will be a provision to download the output in a report form.

Due diligence and KYC.

This section generally deals with selecting out the various user account which does not fully comply with the KYC and due diligence provided, based on the inputs given, normally user account, pan, and aadhar details. The main features are:

  • The AccountUser can upload the list of users and their corresponding KYC details in a bulk format, commonly in CSV.
  • The system will validate the inputs given to comply with the KYC and flag the account that does not complete the KYC details
  • The AccountUser will get a graphical and well as statistical analysis of the output.
  • The report can be downloaded as per the demand of the account user

Palm oil industry

This section generally deals with selecting the areas in Malaysia which are related to the palm oil industry and the employee details. The main features are:

  • The AccountUser can upload the CSV file containing the location, type of trafficking reported if any, the internal/external survey or report scores, etc.
  • The system will validate the inputs given to generate a data map of the location based on the coordinates.
  • The AccountUser will get a graphical and well as statistical analysis of the output.
  • The report can be downloaded as per the demand of the account user.
  • The ability for users to modify the visualization using KQL to customize it to the current requirement

Account security:

Each user is required to log in to the Kibana handle inside the application ensuring that either the user who has created the visualization in the Elastic cloud or the users who are given access to the owner user's organization can view the confidential data.

Future Scope

  • An ML algorithm can be used to create a more efficient model
  • As per the demand of the user, more quality life improvements can be done to the entire web app.
  • The data storage can be migrated to a more robust and premium data center for more security.
  • Using REST APIs to collect data from end-user to create data pipelines to overwrite existing dashboards, so that a single visualization can be re-used multiple times.

About

International Fraud Group Hackathon

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • JavaScript 36.5%
  • PHP 30.4%
  • CSS 22.3%
  • HTML 5.3%
  • SCSS 2.9%
  • Less 2.6%