In order to continue providing relevant content,
the Constitution Club needs the financial support of its membership. 
A $5.00 voluntary monthly contribution is requested.

The AmericaAgain! Indictment Engine will be a mobile app that will be backed by the next generation of machine intelligence technologies. It will be a long-term project that will be refined over many iterations. The general theme is that individual contributions of information will be evaluated, scored and ranked according to algorithms which will profile each public servant and identify issues which may eventually lead to indictment under state laws. A machine learning algorithm is taught to recognize patterns in data using the scientific method: The experimenter creates a hypothesis, selects algorithms to test the hypothesis against the data, observes the results of the experiment, and refines the next hypothesis. We are in the very early stages of envisioning the requirements for this application, so feel free to add comments to this topic if you are motivated to contribute. The video below will give you a sense of the current technologies that may be used in the future Indictment Engine when they mature:

Overview of Microsoft Machine Learning Technologies used in the 201...

After watching the above video, if you want to dive deeper into the types of machine learning experiments that can be done, I recommend this video:

Seth Juarez – Machine Learning for Developers – December 2016

Securing a Distributed Legislature with Bitcoin Technology:

How will a distributed legislature that communicates via public networks ensure that voting on legislation is tamperproof?
The same technology that validates Bitcoin transactions has many non-financial applications for securing and validating distributed transactions.
The security is based on public key cryptography, where the private key is held only by the authentic party of a transaction.
The public key is mathematically derived from the private key and used to derive an address of the party to a transaction.
A network of bookkeeping machines validates all transactions between parties in a manner where fraud can easily be detected because they are mathematically dependent on results from other machines in the network. The concept of Proof-Of-Work can ensure that a bookkeeping machine actually performed the calculations required to validate a transaction.
All transactions are linked together in a “blockchain”. It is this blockchain technology that underlies Bitcoin and related applications.
We would need military grade validation of any proposed solution built to secure voting integrity for a distributed legislature.
Congresspeople would need to be trained in the secure storage and backup of their private keys – and there are related “wallet” technologies that can do this and verify the integrity of wallets.
The fascinating aspect of this technology is that untrusted machines are actually providing high security and transaction integrity.
Welcome to the future. May you live in interesting times.

Views: 127


You need to be a member of Constitution Club to add comments!

Join Constitution Club

Comment by Eric Rangell on March 5, 2017 at 9:03am

An example of voting using blockchain technology:

- A central House server generates individual ballots for each representative, signed with the server's public key

- A transaction sends each ballot to the voting wallet for each representative

- The representative votes by sending the ballot to a "ballot box" entity on the network for the bill being voted on.  Separate ballot boxes exist for each vote: ex: HR899YEA, HR899NAY

- The ballot box can verify that the ballot was issued by the official House ballot server, and was sent from the representative's wallet.  An immutable record of every transaction can be verified by anyone to ensure that the vote is valid.

Comment by Eric Rangell on March 3, 2017 at 9:57am

Play with this page - type some text and see how well the computer understands your sentiment:

Comment by Eric Rangell on March 3, 2017 at 9:54am
Comment by Eric Rangell on February 21, 2017 at 2:32am
Take a look under the hood to learn what powers the Azure Machine Learning engine:
Comment by Eric Rangell on February 20, 2017 at 10:30am

The indictment engine will use the next generation machine learning technologies (which are emerging now and will be perfected in a few years). Here is an example of one - conceptual graph:

If we start with a blank slate of this technology, we would feed it text of law codes and legal decisions and have it infer precedents and patterns that are likely to lead to indictments on State laws. We're just dreaming this up now but it will become more practical as we follow the research.

Also look at this article under the section: "Helping with decision making"

It describes a new service that could potentially be used to teach the Indictment Engine how to make decisions on target potential based on evidence:

Comment by Eric Rangell on February 18, 2017 at 1:38pm

This radio show has an interesting discussion of Article 1 Section 9 (emoluments). We will need to understand the legal technicalities of what is indictable. For example, the burden of proof is on the prosecutor to prove that Hillary Clinton had any influence on the Clinton foundation if she claims that it was run at arms length.

Source: We The People: The Constitution Matters with Gary Porter
on WFYL 1180AM Valley Forge 17 Feb 2017.

Comment by Eric Rangell on February 18, 2017 at 2:23am
The R language has powerful statistical data analysis features. It is now bundled in SQL Server 2016. This Microsoft blog post is a good starting point:
Comment by Eric Rangell on February 17, 2017 at 4:01am

Some notes about how the Indictment Engine may work.

  • Look into Domain Driven Design methodology for analysis of the problem space and creation of a common language for referring to components of the system.
  • AmericaAgain! members would submit Investigation Requests or FOIA requests to be reviewed by the Liberty Labs team.  They fill out a form explaining why the request may be relevant, the parties that may be involved, and submit contact information for follow-up.  Evidence may be submitted as attachments to the request if desired.
  • Requests are pooled and a Kanban system is used to assign follow-up tasks to Liberty Labs team members.
  • The follow-up task involves reviewing the submission, evaluating it, and prioritizing it.  A response is then sent to the submitter indicating the review results and status of their request.  If the submitter desires an appointment to discuss the request with the reviewer, they may schedule it online or by phone.
  • The reviewer fills out an intake form containing the information from the request and information obtained from follow-up interviews with the requestor.
  • Intakes are submitted into the Kanban system for peer review to determine investigation priority and evaluation of evidence.
  • Feedback from peer reviews is kept along with all previous documentation in a dossier.  Evidence is scored and either approved or discarded.
  • FOIA requests are consolidated by a team which prioritizes them and determines an optimal submission strategy.  The status of each FOIA request is tracked along with follow-up correspondence by the FOIA team.
  • FOIA documentation is attached to dossiers when received.
  • When a dossier has enough evidence collected, it is peer reviewed and scores are assigned by each peer across various categories that indicate likelihood of successful indictment.  These numerical scores are fed into Machine Learning experiments in order to compare dossiers with historical results and refine the dossier.  Classification of indictment candidates is done based on results of experiments.  The parameters used for each experiment are archived for audit purposes.
  • A database of State Laws are maintained, and a cross reference to Federal Laws is maintained by the legal team.  As dossiers are reviewed, the database is consulted to determine State Laws which are relevant to the indictment.
  • As dossiers complete the evaluation process, they are presented for scoring and voting by the legal team.  All voting is done using blockchain technology to create a permanent record of decisions.
  • When target dossiers are selected, the investigative team begins publishing blog articles about activities being initiated on the target.  Registered users of the Indictment Engine may read these blog articles and submit feedback, which is moderated.
  • Grand Jury training is held for people interested in serving on specific cases.

To help people understand how the indictment engine may be used, test cases can be created showing how Hillary Clinton or Barack Obama would be indicted on the state laws of their State of residence.  For a catalog of Barack Obama's sins, listen to this show:

Comment by Eric Rangell on February 15, 2017 at 2:43pm
This page will give you an idea of Microsoft's offerings for text analysis:

Machine learning is the engine which powers the Cortana Intelligence Suite.
Your architecture skills will be valuable in designing and evaluating solutions and platforms so we build on a solid foundation.
Comment by Eric Rangell on February 12, 2017 at 10:20am
Here's an interesting website that tracks "spending" per representative based on their votes on legislation. Play around with it and add comments about the user interface and usefulness of information.

The AmericaAgain! Indictment Engine will have multiple dimensions of evaluation of representatives, along with evaluation of evidence submitted by users, which feeds into a scoring engine to identify candidates for indictment.

© 2017   Created by Keith Broaders.   Powered by

Badges  |  Report an Issue  |  Terms of Service