CORDIS oferuje możliwość skorzystania z odnośników do publicznie dostępnych publikacji i rezultatów projektów realizowanych w ramach programów ramowych HORYZONT.
Odnośniki do rezultatów i publikacji związanych z poszczególnymi projektami 7PR, a także odnośniki do niektórych konkretnych kategorii wyników, takich jak zbiory danych i oprogramowanie, są dynamicznie pobierane z systemu OpenAIRE .
Rezultaty
This represents a final version of D2.2 including findings learned from carrying out the co-creation activities.
Preliminary report on co-creation methodologies and findings (odnośnik otworzy się w nowym oknie)Within the first phase of the BIAS co-creation process, every partner will carry out 2 workshops with 35 stakeholders (n ≥ 490 participants)— AI specialists, researchers, students, practitioners, HR specialists, workers and applicants dealing with AI solutions coming from the academia and the private/industry sector—belonging to the national labs. Stakeholders will be the same in both workshops, to the greatest extent possible. The timeframe between the two workshops will be agreed with the WP3 leader. The results of the workshops will flow into instant-reporting that will be elaborated by the facilitators using the template made available by SVEN to feed into the respective tasks in WP3 and the international co-creation workshop, which will consist of the consortium and 1–2 stakeholders per partner. Results for this phase will inform the report on co-creation activities.A preliminary report detailing the co-creation methodologies will be due M12. A final report detailing complete findings will be due M15.
Report on dissemination activities (odnośnik otworzy się w nowym oknie)This will offer a comprehensive report of BIAS dissemination and communication activities including: Dissemination in web and social networks, Distribution of project newsletters and press release, Participation in external events, and publications both scientific and general public. A first report will be prepared by M24, and an updated report prepared in M48.
Preliminary Holistic report on mapping, survey, interviews (odnośnik otworzy się w nowym oknie)This report will consolidate findings from Task 2.1: Mapping current AI use in labor management, gender/diversity needs assessment, Task 2.2: Survey of workers on bias experiences (n ≥ 4000) and Task 2.3: Needs assessment expert interviews (n ≥ 70). These findings will inform the creation and implementation of co-creation activities.Task 2.1: We will map current AI in labor management particularly focusing on gender and diversity issues. We will focus on a two-fold approach in which we will first review different applications found in the market on AI for work management to find out how they consider diversity and inclusion aspects, including from the design, implementation, and execution of the tasks.Task 2.2: We will conduct an online survey geared informed by preliminary results from T2.1 toward understanding how workers react to and make sense of algorithmic profiling when it is made visible to detect and understand their perceptions of existing biases. The methodology will rely on qualitative accounts as well as quantitative data from a European representative sample of at least 4,000 respondents from all 27 EU member states as well as Iceland, Norway, Switzerland, and Turkey.Task 2.3:This task will consult with experts—AI developers and HR executives—regarding current practices concerning AI in recruitment and HR practices and what concerns they have regarding bias. A short, easy to follow interview guide supporting interviews of approximately 30 minutes will be created featuring questions regarding what the experts view to be the most important issues with multiple choice responses allowing for further elaboration through open conversation.I perliminary, but mostly complete report will be prepared by M9 that can inform the co-creation activities. In the event that it takes more time than envisioned to gather 4,000 survey responses, a final report will be delivered in M15.
Preliminary report of raising awareness activities, including findings and guidelines for holding similar activities (odnośnik otworzy się w nowym oknie)Raising awareness activities concern: 1. The development and dissemination through theproject’s communication channels (WP7) of mini-videos about what constitutes bias, how they are experienced and how to recognize them. The videos will be developed with the support of LOBA as WP7 leader and BFH as WP3 leader. They will be disseminated to constituents in line with our communication & dissemination plan, with a focus on citizen advocacy organizations and educators. 2. Online raising awareness sessions: the sessions will be organized in the form of webinars by the task leader and held in English. Three rounds of raising awareness sessions will be organized (in M12, M24, M36). The sessions will focus on the importance of tackling gender and intersectional biases in AI and on how to prevent such biases.The content will be less specialized than the T5.3 and T5.4 events and promotion will focus on the same constituencies of the informational videos. They will be organised with the support of SVEN. The activities will be evaluated through questionnaires. The activities and evaluations will be reported on in this deliverable.A preliminary version of this report will be due M30. It will contain information about the scripts used and awareness raising sessions carried out up to that point. A final report will be due M46 that will detail all sessions as well as learnings from evaluations and assessments.
Plans for provision for open source software (odnośnik otworzy się w nowym oknie)BIAS intends to make software developed during the project available open-source. A public deliverable describing these plans will be created to complement the confidential business plan. A perliminary plan will be delivered M24 and a final report on what open-source software was made available and where it can be found will be delivered in M48.
Capacity building program, scripts, material (odnośnik otworzy się w nowym oknie)Different capacity building programs tailored to the different participants’ profiles (researchers, students, AI practitioners, HR specialists, employees and eventually others identified in T5.1) will be designed. AI researchers, students, practitioners, HR specialists, workers and applicants dealing with AI solutions coming from the academia and private/industry sector at each partner’s county level will represent the main target groups. As far as the formats are concerned, both an online and an in-person format will be designed to easily adapt to the unpredictable constraints related to the current pandemic situation. While the in-person format will be constituted by 3 half day sessions, the online one will be scheduled in 3 different sessions of 3 hours each. An interactive and hands-on learning approach will be ensured through the adoption of participatory techniques and methodologies (e.g. group works, interactive and participatory tools, etc.). This deliverable will collect the programs, scripts, and materials developed during the first iteration of the capacity building program.
Final report on interviews conducted and preliminary research findings (odnośnik otworzy się w nowym oknie)This will be a final version of D4.1. All ethnographic fieldwork will be completed by M30, and this deliverable will provide details on all fieldwork conducted and any preliminary research findings. A full report on ethnographic findings will be delivered in D4.3 in M42.
Interview guide and data on interviews conducted (odnośnik otworzy się w nowym oknie)The ethnographic fieldwork of BIAS will consist of qualitative interviews with 365 informants throughout Europe: (1) developers of technology that has been used in labor settings (n ≥ 50); (2) employers and HR personnel (n ≥ 90); (3) employees and trade union representatives (n ≥ 225). This report will detail the interview guide used for each type of interview and provide anonymised demographic information concerning the informants.A preliminary report will be due M12. This will include the interview guide, report on interviews conducted thus far, and plans to conduct the remaining interviews. A final report detailing information on all interviews conducted will be due M30.
Preliminary report on ALTAI analysis (odnośnik otworzy się w nowym oknie)We will implement the Assessment list of Trustworthy AI (ALTAI) together with the developers and technical parts in our project so we can identify what elements can be implemented in the design, implementation, and execution of the AI systems. The idea is to run the assessment in a life-cycle and co-creational way between the technical and the social scientists of the consortium. This will result in a public report detailing the experiences with the BIAS technology in particular as well as provide general reflections on how the ALTAI assessment can be implemented in practice.A perliminary report on the analysis will be delivered in M30 and a final report in M48.
The Dissemination, Exploitation, Communication plan includes the detailed planning of all communication activities in a systematic manner. This involves:▪ Creation of actions: creative definition of the communication action as well as the briefing of how this will be put into practice.▪ Definition of Objectives: contextualization of the action in terms of how each action/message will be adapted to suit any particular target audience.▪ Definition of and assignment to the responsible party to carry out each action.▪ Timing: definition of the time period of the action based on prior coordination with the other WPs.▪ Materials: definition of the communication materials to be created.A preliminary plan will be delivered in M3, with a revised plan delivered in M24
Revised dissemination exploitation, communications plan (odnośnik otworzy się w nowym oknie)This represents a revision of D7.1.
This is a revised version of D1.2 detailing what data has been gathered and produced thus far in the project and the practices undertaken to manage this data using FAIR principles. It will also make plans for the final implementation period of the project.
Data Management Plan (odnośnik otworzy się w nowym oknie)The Data Management Plan detailing all of the data created by BIAS and how thedata will be protected while also making it as openly accessible as possible according to the FAIR principles (Findable, Accessible, Interoperable, and Reproducible). The DMP will outline how we will decide between making data available and maintaining confidentiality for IPR or privacy concerns. The initial Data Management Plan will be prepared by M6 and it will be revised in M29 and M47 in line with reporting periods.
Publikacje
Autorzy:
Carlotta Rigotti, Eduard Fosch-Villaronga
Opublikowane w:
Computer Law & Security Review, Numer 53, 2025, ISSN 2212-473X
Wydawca:
Elsevier BV
DOI:
10.1016/J.CLSR.2024.105966
Autorzy:
Alexandre Puttick, Carlotta Rigotti, Ahmed Abouzeid, Eduard Fosch-Villaronga, Mascha Kurpicz-Briki, Pinar Øztürk
Opublikowane w:
AIMMES 2025 - AI bias: Measurements, Mitigation, Explanation Strategies, 2025
Wydawca:
CEUR
DOI:
10.5281/ZENODO.15649867
Autorzy:
Leander Rankwiler, Mascha Kurpicz-Briki
Opublikowane w:
Workshop Proceedings, Links currently to pre-print, final link will be added later, 2024
Wydawca:
We have this information yet, there is only a pre-print on the conference website so far.
DOI:
10.5281/ZENODO.15649156
Autorzy:
Catherine Ikae, Mascha Kurpicz-Briki
Opublikowane w:
AIMMES 2025 - AI bias: Measurements, Mitigation, Explanation Strategies Proceedings of the 2nd Workshop on AI bias: Measurements, Mitigation, Explanation Strategies, 2025
Wydawca:
CEUR
DOI:
10.5281/ZENODO.15649549
Autorzy:
Silvia Ecclesia
Opublikowane w:
AIMMES 2025 - AI bias: Measurements, Mitigation, Explanation Strategies Proceedings of the 2nd Workshop on AI bias: Measurements, Mitigation, Explanation Strategies, 2025, ISSN 1613-0073
Wydawca:
CEUR
DOI:
10.5281/ZENODO.15649766
Wyszukiwanie danych OpenAIRE...
Podczas wyszukiwania danych OpenAIRE wystąpił błąd
Brak wyników