How can we harness data and information for the health of communities?
105 contributions
626 ideas
626 final ideas
39 final ideas
39 final ideas
Announced!
Personal Life Planning Assistant
We intend to release an easily installable set of applications for personal computers and phones which compiles health information from trusted news sources into peer-reviewed executable rules that aid in designing and executing a life plan.
Written by Andrew Dougherty
Using various GPLv3 software under continuous development for over 14 years, we intend to release an easily installable set of applications for personal computers and phones which compiles health information from user-specified trusted news sources into peer-reviewed executable rules that aid the user in designing and executing a continuous, social life plan to achieve and maintain health and eudaimonia.
Free and open source software provides an ideal platform to distribute life saving and improving services through the unrestricted redistribution of software that implements critical functionality for users, especially those in poverty and/or having disability and/or illness. The chief difficulty tends to be in funding its development.
We propose the Personal Life Planning Assistant (PLPA) as follows:
News sources and public domain texts are first vetted using Natural Language Processing software (NLP). The process is divided into two steps, the acquisition of knowledge (using existing and custom knowledge extraction and management applications), followed by Rule Extraction algorithms. This creates a library of unrefined rules, which are then refined via a crowd-sourced, crowd-reviewed, trust-managed, prospective and retrospective platform, to create a repository containing life planning, health and wellness rules in various stages of refinement.
In addition to rules, relevant facts about the person's life situation and position, such as life goals and health information, are collected en masse both passively via NLP through social media and submitted resources such as documents, and actively disambiguated through knowledge elicitation dialogs, and are stored locally in a secret secure encrypted knowledge base (KB) (and all effort is put forth to maintain confidentiality, integrity and availability of that KB).
The rules previously collected are to be applied to this collection of facts about the person in question in order to suggest additional facts to add securely to the knowledge base, including information on potentially recommended actions to take in certain circumstances.
The theories of rationality, practical reasoning and multi-agent logics inform the development of argumentation-based practical reasoning software which functions as a decision support system for the user over the collection of personal facts and trusted rules to reason about which of the goals collected and/or elicited from the user are practical and useful to achieve given their life situation. Planning, scheduling and interactive execution is then undertaken to develop conformant plans across multiple levels of granularity which provide options and help guide the user from their phone and computer towards the achievement of their goals.
Several kinds of plans are considered, such as diet, preventative and emergency medical, short, mid and long-term life temporal and resource planning. The user's performance in achieving their goals (health or otherwise) is monitored and the effectiveness of the various plans, and correlations between such factors as diet and health conditions can be data-mined (given the user's fine-grained consent or with secure multi-party computation) to obtain large datasets for evidence-based medicine, public health policy, etc. Public health can then be influenced directly by improving the quality of the public rule sets. These rule sets are then calibrated and tweaked by running the PLPA planner on a regression test library of sample life situations, enabling a scientific operationalist approach to life. Note that, far from running your life, this allows the user to specify their tastes, preferences, trusted sources, opinions, etc, and for plans to be computed on this basis. Also, being open source enables people to modify it to their purpose.
To Summarize:
News sources (news-monitor (+RSS feeds, ~websites)) and CC and public domain texts (corpora-manager (+openlibrary.org, +wikipedia, etc)) are vetted for health, wellness and long-term planning information, which is analyzed using Natural Language Processing (NLP) software (Formalize2, KNext, NLU?, Workhorse) and yet-to-be-obtained rule mining software (such as "A framework for the automatic extraction of rules from online text" and "Extracting STRIPS Representations of Actions and Events") and extended through a crowd sourced, reviewed and trust-managed framework (preferably based on Perl Catalyst)) to create an argumentation-based practical reasoning expert system (Action-Planner, CLIPS, Gamer?, Cadiaplayer-3.0?, Leo-II?, EPR?, CaSAPI); which -- using a knowledge base of encrypted personal facts (FreeKBS2, Vampire-KIF), mined from social media using NLP, interactive dialogs, and observed actions (UniLang) -- then helps the user to generate plans and answers (Action-Planner, Verber, Gamer, LPG-td, SPSE2, QUAC) from the compiled rules, facts and stated goals (SPSE2, POSI, UniLang) to quickly obtain relevant, timely and actionable intelligence and advice, administered through multiple modes (such as a cell-phone based interactive execution monitor (IEM) and question answering assistant (QUAC)) in order to more effectively lead a healthy life in dynamic and possibly resource constrained environments such as is experienced due to poverty and/or illness.
I have been dedicated to this project since first realizing the opportunity provided by free software to fulfill the promise of a social safety net through the power, reliability and communicative capacity of modern computers and free and open source software. The chief benefit of the PLPA is that other perspectives, ideas and strategies (that are present in the entire collection of freely available texts and news sources) the user and their associates may not have known about or considered may be applied immediately, directly, efficiently and confidentially to their exact life situation, vastly improving their objective conditions. The user is thus spared the impracticable necessity of reading and evaluating the entire corpus of health and life planning resources, and the need to tabulate whether all the different advice applies to their ever-changing situation - the process is automated, enabling an abundance of choice and possibility, consistent with their stated purposes and preferences, to which most often the user would otherwise have remained unaware.
Currently we are in the process of releasing critical portions and rewriting and expanding essential parts of the FRDCSA (versions 1.0, 1.1, 1.5 and 2.0), which is the overarching project to which the flagship PLPA project belongs. Most essential hardware infrastructure is in place. Funding will go to hiring developers to assist with the release of the FRDCSA. We already have daily scrums for POSI, which is being integrated into FRDCSA-2.0. That software will be freely available, including the software development and testing infrastructure (Vagrant VMs).
Some other notable systems that would be released and/or improved include the Gourmet Meal Planner, the Paperless-Office and the Job Search system.
References:
Overview:
http://slashdot.org/index2.pl?fhfilter=frdcsa
Internal:
http://frdcsa.org/~andrewdo/WebWiki/FreeLifePlanningCoachSoftware.html
http://frdcsa.org/~andrewdo/writings/Overview.odt
- http://frdcsa.org/~andrewdo/writings/Overview.pdf
- http://frdcsa.org/~andrewdo/writings/Overview.html
http://frdcsa.org/~andrewdo/writings/Temporal-Planning-and-Inferencing-for-Personal-Task-Management-with-SPSE2.pdf
http://frdcsa.org/
http://frdcsa.org/frdcsa
http://frdcsa.org/~andrewdo/writings/flourish-2009.odp
http://frdcsa.org/~andrewdo/WebWiki/UtilityMaximizationSystem.html
http://frdcsa.org/~andrewdo/writings/workhorse.odp
http://frdcsa.org/~andrewdo/writings/frdcsa-2.0-requirements-draft.txt
https://github.com/aindilis/myfrdcsas-doc/blob/master/MyFRDCSAs.README
http://frdcsa.org/frdcsa/internal/akahige/
http://frdcsa.org/frdcsa/internal/gourmet/
http://frdcsa.org/~andrewdo/projects/paperless-office/
http://freecode.com/projects/job-search-rapid-response
http://frdcsa.org/frdcsa/internal/verber/
http://frdcsa.org/frdcsa/internal/formalize/
http://frdcsa.org/~andrewdo/projects/nlu/
http://frdcsa.org/frdcsa/internal/quac/
External:
http://frdcsa.org/~andrewdo/essential-reading/data/ai-research/
http://www.cs.rochester.edu/~schubert/projects/world-knowledge-mining.html
http://www.defeasible.org/ruleml2011ijcai/files/OConnor.pdf
http://www.cis.temple.edu/~yates/papers/2011-ranlp-strips-camera-ready.pdf
http://www.catalystframework.org/
http://page.mi.fu-berlin.de/cbenzmueller/leo/
http://page.mi.fu-berlin.de/cbenzmueller/papers/J25.pdf
http://www.wietskevisser.nl/research/epr/
http://www.doc.ic.ac.uk/~ft/CaSAPI/
http://www.aaai.org/ojs/index.php/aimagazine/article/view/2427
http://cadia.ru.is/wiki/public:cadiaplayer:main
http://link.springer.com/article/10.1007/s13218-010-0078-3
http://aaai.org/ocs/index.php/ICAPS/ICAPS11/paper/download/2699/3159
http://zeus.ing.unibs.it/lpg/
http://openlibrary.org
Describe your project in one sentence.
The FRDCSA helps to improve lives by using free software social artificial intelligence to freely distribute software solutions to difficult problems that affect people.Who is the audience for this project? How does it meet their needs?
Anyone with a computer or a phone. It meets their needs by assessing them, strategizing with them, incorporating advice from freely available sources, and aiding them in designing and executing complex plans to resolve difficult situations and facilitate eudaimonia.What does success look like?
Software released and available for all platforms, informing people's choices, improving their lives, especially the lives of those in poverty and/or with disabilites and/or illnesses, and assisting interested parties in redistributing the infrastructure to others who express interest, regardless of on which side of the digital divide they reside..Your Location
Aurora, IL 60506, USA
0 comments
Join the conversation:
Comment