Grants:IEG/Wikintelligence

From Meta, a Wikimedia project coordination wiki
Jump to: navigation, search

status: withdrawn

Individual Engagement Grants
Individual Engagement Grants
Review grant submissions
review
grant submissions
Visit IdeaLab submissions
visit
IdeaLab submissions
eligibility and selection criteria

project:

Wikintelligence


project contact:

Edgar Hagenbichler

participants:



grantees: Edgar Hagenbichler


summary:

Feasibility study by a scientific research institute, for using Wikidata as open Artificial Intelligence for improving quality of Wikipedia articles

engagement target:

Wikidata, pywikipediabot-framework, Commons and Wikipedia (German, English, Simple English, Spanish)

strategic priority:

Improving Quality and Increasing Participation

total amount requested:

16.000 EUR


2013 round 2

NOTE: This proposal is withdrawn due to a delay in our plans of applying for a grant with FFG and due to the results of the discussion of our proposal. We plan to make a new proposal based on the lessons learned; e.g. participation of the involved communities and clarification of the goals.

THANK YOU VERY MUCH FOR DISCUSSING THE PROPOSAL!

NOTE: This proposal was first released as a WMF grant submission of an individual under Grants:Projekt_ANA/Wikintelligence. Now it seems to me that it would better fit to the IEG-Individual Engagement Grants, that is why it is proposed here. It only differs in the project roadmap since the IEG project has to be made within 6 months. Sorry for the inconvenience of reviewing two proposals for the same project.

Project idea[edit]

Project goal[edit]

This statement describes what will be accomplished if the project is successful.
There will be:
  • in 2014 the feasibility study of the scientific research institute will be published under CC-BY-SA 3.0 on Meta-Wiki. The feasibility study shows how WMF can reach most effectively and effeciently the goal of Wikintelligence - OpenAI within the next twelve years.
    • An exemplifying prototype - as part of the feasibility study - within the pywikipediabot framework and within a virtual robot platform (e.g. SmartBody) with the use of Wikidata will be released in 2014.
    • A roadmap until a year in the future, probably within the next 10 -15 years, and potential funding organizations will be evaluated as part of the feasibility study. As "longterm-goal 2025" Wikintelligence - OpenAI shall improve Wikipedia articles and shall provide a dialogue system with natural language using a real humanoid robot. It shall pass the Total Turing Test in the Loebner contest in a year in the future, probably within the next 10 -15 years,
      • The roadmap that will be evaluated in the project includes:
        • Roadmap-subgoal 1: Wikintelligence will expanded Wikidata with human-neurophysiological properties and will integrate an open source virtual robot platform until 2016.
        • Roadmap-subgoal 2: Wikintelligence will develop and use rule-based clustering data classification techniques for usage on Wikidata within the Pywikipediabot-framework until 2016.
        • Roadmap-subgoal 3: Until 2016 the Wikintelligence project members will develop performance criterions, show cases and use cases within Wikidata and Wikipedia. This will include 1. detecting and defending spam attacks, 2. detecting of inconsistencies between different language versions of an Wikipedia article, 3. checking currentness of data and references of an Wikipedia article and (semi-)automatic updating, 4. (semi-)automatic categorizing of Wikimedia Commons media, 5. (semi-)automatic update of help pages of Wikidata.
        • Roadmap-subgoal 4: Wikintelligence will pass the "weak" Turing-test, defined as "chatting like a human" until a year in the future, probably within the next 5-10 years.
        • Roadmap-subgoal 5: Within the Wikintelligence-project we will finish (another) feasibility study how to pass the Total-Turing-test until a year in the future, probably within the next 5-10 years, defined as "videochatting like a human until a year in the future, probably within the next 10 -15 years"
      • The potential (partly-)funding organizations that will be evaluated in this project include:
        • Wikimedia Foundation
        • FFG - Austrian Research Promotion Agency
        • The Neurorobotics Platform of The HumanBrain Project of the 7th framework programme of the EU, ramp-up phase starting in April 2014, lasting for two years
        • The Medical Informatics Platform of The HumanBrain Project of the 7th framework programme of the EU, ramp-up phase starting in April 2014, lasting for two years
        • Austrian Federal Ministry of Health
        • Main Association of Austrian Social Security Institutions

Part 2: The Project Plan[edit]

Project plan[edit]

NOTE: This proposal was first released as a WMF grant submission of an individual under Grants:Projekt_ANA/Wikintelligence. Now it seems to me that it would better fit to the IEG-Individual Engagement Grants, that is why it is proposed here. It only differs in the project roadmap since the IEG project has to be made within 6 months. Sorry for the inconvenience of reviewing two proposals for the same project.

Scope:[edit]

Over the course of 6 months, we want to make a feasibility study by a scientific research institute how to provide a way for realizing an open Artificial Intelligence based on Wikidata, Pywikipediabot and a vrtual robot platform. Long term goal is, to improve the quality of Wikipedia and to support free access to knowledge with an intelligent dialogue system and automatic contributing system in the future, probably during the next 10-15 years.

Scope and activities[edit]

The project designer has already ordered at OFAI a so-called Innovation Voucher, funded by the FFG - Austrian Research Promotion Agency. As a positive result of this the OFAI recommended to perform a feasibility study.

The FFG offers as a next step the funding of 75 % of the cost of a feasibility study, this means maximum 30.000 Euro with the cost of 40.000 for the feasibility study. The cost for the project designer are 5.000 Euro plus 20 % VAT and are not included and will not be funded by the FFG. FFG will decide about funding on 4 December, 2013. Therefore the application together with an offer of OFAI about the content and the methods of the feasibility study has to be at FFG 6 weeks prior - i.e. until 21 October, 2013.

If FFG and WMF both will fund the feasibility study, then I will order the study from OFAI (probably on 5 December, 2013). A kick-off workshop is planned in January 2014 at the University of Gran Canaria. Monthly meetings are planned with the members of OFAI and me to conduct the study. The study finishes at 1 July, 2014.

Tools, technologies, and techniques[edit]

"Feasibility study for Wikintelligence" project falls into the Outreach program area since it will support Wikimedians to build relationships with scientific and educational institutions, in particular with OFAI - Austrian Research Institute for Artificial Intelligence, Medical University of Vienna and The University Institute of Cybernetic Science and Technology of Gran Canaria as well as with local authorities as the Federal Ministry of Health or the Main Association of Austrian Social Security Institutions. A prototype as part of the feasibility study within the pywikipediabot framework and within a virtual robot platform (SmartBody) with the use of Wikidata will be released.

Budget:[edit]

Total amount requested[edit]

Total project budget

55.000 EUR

Total amount requested

16.000 EUR

Are there additional sources of revenue that will fund any part of this project? List them here.
  • FFG: 30.000 EUR
  • Edgar Hagenbichler: 9.000 EUR (4.000 EUR VAT 10% that I can count up for the local tax authorities for the OFAI-study plus 5.000 EUR as inhouse-effort for 62,5 hours of project design)

Budget breakdown[edit]

Number Category Item description Unit Number of units Cost per unit Total cost Currency Notes
1 work hours scientific researcher OFAI for Feasibility study - 1 hour 1 hour 1.000,00 40,00 40.000,00 EUR
2 VAT 10 % VAT 10 % OFAI - 1 hour 1 hour-VAT 1.000,00 4,00 4.000,00 EUR
3 work hours project designer - Edgar Hagenbichler - 1 hour 1 hour 125,00 80,00 10.000,00 EUR
4 VAT 20 % VAT 20 % Hagenbichler - 1 hour 1 hour-VAT 62,50 16,00 1.000,00 EUR

Non-financial requirements[edit]

We will need for team building and for the workshops 10 Wikipedia T-shirts, 10 caps and 10 coffee cups. Furthermore we want to use the Meta-wiki logo, the Wikipedia logo, the Commons logo and the Wikidata logo for the report of the feasibility study and for the workshops.

Intended impact:[edit]

Target audience[edit]

This project will directly benefit Wikidata and the pywikipediabot-framework, as we expect to provide an exemplary prototype as a result of the feasibility study.

Fit with strategy[edit]

Currently Wikidata is financed with donations of the Allen Institute for Artificial Intelligence, of the Gordon and Betty Moore Foundation and of Google Inc. with a total of 1.300.000 Euro. IBM Research’s Watson team donated its 10.000 US$ 2013 Feigenbaum Prize of the Association for the Advancement of Artificial Intelligence (AAAI) to Wikidata.

The present project "feasibility study for Wikintelligence" will provide input on the long-term goals of the Wikimedia community projects. It will increase Reach, Participation and Credibility by engaging a scientific research institute with very high reputation and impact on the scientific community to work on Wikidata. Organizational maturity and effectiveness and financial sustainability will be supported by offering new ways of co-funding Wikimedia projects by official authorities. It will spread the mission to collect and develop the world's knowledge and to make it available to everyone for free.

Sustainability[edit]

This project has a high potential to be replicated by other groups in the European region. For one thing, we will publish application for grants (if not prohibited by non-disclosure stipulations by the potential grant-giver) and will be thoroughly documenting the successes and failures of this project. This can be used as a project manual for other organizations and groups doing similar work in local region. You can find the application for the Innovation Voucher and the result of OFAI here. Since the feasibility study will show a way how to reach Wikintelligence - Open Artificial Intelligence it will have impact on funding of further development of the pywikipediabot-framework and for producing manuals, including e.g. follow-up projects of Wikidata Toolkit or Understanding Wikidata.

Measures of success[edit]

We will consider this project a success if
  • the feasibility study is successfully conducted by a scientific research institute, namely by the OFAI - Austrian Research Institute for Artificial Intelligence and is published in Meta-wiki under a free license
  • a prototype as part of the feasibility study within the pywikipediabot framework and within a virtual robot platform (SmartBody) with the use of Wikidata will be released under the appropriate free license (http://tools.wmflabs.org/pywikibot/core/LICENSE).
  • a kick-off has happened with two workshops (1 week each) in Gran Canaria, Spain: Together with The University Institute of Cybernetic Science and Technology and/or Innovation Centre for the Information Society (Centre of R&D+i (Research, Development and Innovation)) of the University of Las Palmas de Gran Canaria and with OFAI.

Participant(s)[edit]

Discussion[edit]

Community Notification:[edit]

Please paste a link to where the relevant communities have been notified of this proposal, and to any other relevant community discussions, here.

Endorsements:[edit]

Do you think this project should be selected for an Individual Engagement Grant? Please add your name and rationale for endorsing this project in the list below. Other feedback, questions or concerns from community members are also highly valued, but please post them on the talk page of this proposal.

  • Community member: add your name and rationale here.
  • It is not entirely clear to me what this project is about, but it sounds like the people involved are very capable. Funding some basic research on how to handle and update the unimaginable large data volumes of all Wikimedia projects can only be of benefit. Plus there seems to be an already ongoing multiplication factor, where the people asking for the grant are getting their students and colleagues involved. --Tobias1984 (talk) 15:59, 31 October 2013 (UTC)