Skip to main content

Revealing fair and actionable knowledge from data to support women’s inclusion in transport systems

Periodic Reporting for period 1 - DIAMOND (Revealing fair and actionable knowledge from data to support women’s inclusion in transport systems)

Reporting period: 2018-11-01 to 2020-04-30

Current transport systems do not sufficiently take into account physical and social characteristics of women in the design of products and services, and in fostering women’s employability in the industry.
Many women, due to socially constructed norms, experience transport differently. They have different mobility patterns and their perception and use of transport in general and public transport is different from men:
- In the family unit, women usually take care more of the children or the elderly than men, what directly affects the employment conditions: women have more part-time work and/or staggered schedule than men.
- Women have a trip-chain pattern of mobility, introducing usually care and household related mobility aspects (i.e. picking or leaving the children at school or food purchase) men tend to move from A to B. In addition, when using public transport, they have to face some infrastructure mobility barriers related to the use of strollers, shopping carts, or carrying children.
- Also, women usually feel less safe than men when using public transport, since they are more likely to experience harassment or undesired behaviours by other users of the public transport system.
- Less than a 20% of women work in the transport sector.
Therefore, it is important to the society to improve the experience of women as users of the transport system and to increase the recruitment and inclusion of women in transport related companies.
DIAMOND’s main goal is to turn data into actionable knowledge with notions of fairness, in order to progress towards an inclusive and efficient transport system. Identifying, designing and evaluating specific measures for fulfilling the needs and expectations of women as users of different transport modes and as jobholders in the sector. The knowledge gathered in the data analysis will be analysed by interdisciplinary methods from social sciences and computer science to provide fairness measures. These results will then be fed into a toolbox to provide recommendations on how to achieve fair inclusiveness for women in different scenarios.
To proof actionability, this project will make concrete advances in four real-world scenarios (use-cases) where inclusiveness is currently a central issue: 1.- railways and public transport, 2.- Autonomous vehicles, 3.- bicycle sharing and 4.- employment.
DIAMOND arrives to its mid-term with sound progress and all activities going according to the plan. During this period, DIAMOND has concreted a conceptual framework and a methodology capable to achieve outputs and outcomes for different kind of women and capable to quantify the validity of these outcomes and results .Main progress in this first period is arriving to a definition of “woman” for a good delimitation of the project scope. The definition of “women”, based on the definition of gender as a social construction which changes over time and space, was deeply analysed and defined for a better definition of the target which project’s outputs and outcomes should be addressed to.
The concept of Polyhedral Individual (PI) was analysed and concreted in the most relevant characteristics that could potentially act as discrimination vectors (age, religion, cultural level, sexual orientation, etc...) defining ranges per each of them.
These ranges were defined based on the data available regarding the population size in Eurostat and other sources for some PI’s characteristics, and for others based on the option of being able to trigger different preferences among women. This work of definition of PI’s characteristics and ranges was carried out in cooperation with TInnGO, our sister project, in order to allow a potential future comparison of results.
DIAMOND has also progressed in the fairness concept applied to the transport system and to each of the use cases addressed by the project, due to fairness is a very general concept. What fairness means per each use case, how to generate fairness measures and how to assess them, are questions that DIAMOND has faced on during all this period. In parallel, a methodology to generate and assess fairness measures have been defined per each use case.
The generation of fairness measures is based on a previous identification and hierarchy of fairness characteristics (FCs), or relevant criteria with influence on the equity of the transport system, and the assessment of fairness measures is based on the maturity model and the Rasch model in order to quantify how each fairness measure is matching needs of users and employees of the transport system.
A hierarchical model of FCs in 3 levels was established by DIAMOND, and two kind of data is being collected for each of these FCs: Data regarding users and employees’ perception and Data regarding service providers and employers.
During this period, an interim data analysis with the available data was conducted by applying the AHP (Analytic Hierarchy Process) and factor analysis and regression models. In both cases, algorithms are ready to be repeated on a continuous basis as we get more and more data.
Current results are on one hand a weighted hierarchy of fairness characteristics (FCs) level 1 and 2 per specific profiles built on the PI’s characteristics and ranges, which allow the generation and assessment of fairness measures through the Maturity assessment and Rasch models, and on the other hand the identification of more sensible profiles regarding the inclusiveness in transport per each use case.
DIAMOND methodology is specific to analyse the transport sector from all the possible perspectives and for all kind of profiles thanks to the Inclusion Diamond and the Polyhedral Individual approach, which have been taken into account along all the study development during this periodic report.
The analysis of data is based on real data obtained from databases, observations and Social Media, and their treatment using Multicriteria Decision Methods as the AHP (Analytic Hierarchy Process), providing the study with a holistic approach, and machine learning algorithms as factor and regression analysis and Bayesian Networks. After this an interdisciplinary analysis with sociologists, psychologists, technicians, transport experts, etc. is applied for the fairness measures generation and the design of the fairness assessment model through the Maturity and the Rasch model.
DIAMOND will structure the study in an organised way through the Polyhedral Individual and diamond representation. The combination of a multi-level interdisciplinary analysis together with the development of a self-diagnosis tool and a decision support system is enabling a high-level understanding of women’s needs according to their specific characteristics. This methodology is providing specific outcomes and results per profiles, providing fairness measures and fairness assessments for specific granularities of women, also considering specific value of validity for these results dependent on the sample size and the strength of preferences showed in the surveys. Previous gender studies, even those applied to the transport, are not structured in a so technical and methodological way. This is a progress in the gender analysis field, and the methodology prepared by DIAMOND will be by contribute to further studies addressing other use cases and specific women profiles.
diamond-model.png