(Dept. A – Principal Investigator[2])

D7. Greek Politics

(max. 24 months)

(max. 30.000€)

The proposed project can be considered as a first step towards the goal of developing professional forecasting models to address similar problems and it is entirely interdisciplinary as it requires a continuous synergy between scientists of different origin and research areas, who, working as a team, will make their research tools available for the treatment of the problem.  In particular, the project involves two Departments of the University of Crete with a long record of research achievements related to the subject of the proposal: The Department of Political Sciences with its long experience on migration issues will collaborate with the Department of Mathematics and Applied Mathematics which has long research experience in optimization problems using modern computational tools. Political Scientists will describe the problem and perform the necessary qualitative analysis of the available data, helping the Mathematicians to construct their prediction tools and train their models to provide a reliable estimation of future trends. It is obvious that a single team only could not achieve the objectives of the project! Noticeable that it will be the first time that the two Departments will join their forces for treating such a critical problem.

Research Issue and Conceptual History: Taking as a starting point the unpreparedness of Greece to manage huge migration/refugee flows during the refugee crisis of 2015 and so on, one cannot but point out the fact that the chronic inconsistency of migration policy making in Greece, combined with the sudden influx of migrants/refugees in the landmark year of 2015, revealed the element of unpredictability of massive migrations flows and the disruption they cause in host countries during its occurrence, and in this case, Greece. As such, the unpredictability element is what this project aspires to tackle. By using combined insights from the disciplines of political science and mathematics, this projects aims at the construction of a start-up pilot prediction model to use in the Greek case for the short and long term based on three different flows scenarios: hard, medium, soft. WHY? Evidence shows that during the refugee crisis of 2015 and the period it followed it, a new migration policy began to emerge In Greece. This policy began to systematically transform the core of Greece’s migration landscape leading to mostly fragmented attempts to manage migration in the context of the so-called five nodes of the transformation of the Greek Migration Policy, namely, the European Migration Agenda signed in May 2015, the EU-Turkey Joint Statement of 18 March 2016, the Instrumentalization of the Refugee Issue by Turkey in February 2020, the arrival of the Corona Virus Pandemic (Covid-19) and the New Pact on Migration and Asylum signed in September 2020 (Papadakis & Dimari, 2021 under publication). Despite the novelties inserted in the Greek migration policy corpus in the context of the five nodes, yet, the challenges remain, and the necessity for the formulation of a viable migration policy is becoming a requirement. This became quite evident during the Evros February crisis in 2020 whereby an estimated 13,000 refugees/migrants found themselves in Evros, trying to enter Europe. The result was thousands of people being trapped between the borders of Greece and Turkey, causing a huge crisis between the two countries and revealing Turkey’s intention to use the refugee issue to its benefit (Lappas, 2020).

Research Questions: But how can we approach the construction of a pilot prediction model of short and long term migration flows that will be based upon three different flows scenarios that is hard, medium and soft? In order to do so, the two questions to answer are what data do we need? In addition, what are the flows’ scenarios going to entail? As far as the first question is concerned, we are going to use configured quantitative data on migrant/refugee flows’ intensity and decrease in Greece from the 1990s, when Greece switched from a sending country to a receiving one, up to 2021, when the New Pact on Migration and Asylum was signed. Regarding the second question and as previously stated, hard flows scenarios pertain to events such as wars and catastrophic events that could trigger massive migration and refugee flows in Greece. The foci here are the Middle East and Africa. The Medium flows pertain to manageable numbers of migration whereby there exists certain ‘normality’ in the arrival of migrants that come in Greece mainly for job reasons. Last, the soft flows scenario pertains to events that trigger the decrease of migrants/refugees flows in Greece, such as for instance, pandemics.

Quality and Credibility: The quality/credibility of the proposed project lie in its necessity. The domino effect of huge migration flows in Greece as a forefront country during the refugee crisis of 2015 has disrupted the already fragmented attempts to manage migration and has revealed the non-resilience of the Greek state mechanism when faced with huge and unpredictable migrations flows. Indeed, the intensity, the urgency and the unstoppable nature of the migration phenomenon/refugee crisis when combined with the element of unpredictability, leave migration policy makers at limbo as any attempt to draw a sustainable and long term is subjected into the unpredictability of migration flows component. As such, this is the challenge this project aspires to tackle: to construct a start-up pilot migration flows’ prediction model that will essentially assist migration policy makers to draw management of migration strategies dependent on relevant flows scenarios (hard, medium, soft scenario). The quality and credibility of the proposed project also lie in the capacity of this project to give form to a novel, pilot prediction model that will act as a living organism and will have the capacity to generate new data that will lead to new policy scenarios according to the data that it will be ‘fed’ with. As such, it is expected to pave the way for further research and tangible policy interventions pertaining to the management of migration in Greece at a long term. The proposed research is promising as it has not been yet approached in the Greek case (see section 2.2), thus any attempt to do so will lead to the mapping of future migration flows and will also help shape a management migration strategy that could be a derivative and a synthesis of each of the two departments involved scientific data. This tool will serve the purpose of a broader strategic reflection on future migration flows and their possible drivers in Greece and will be the first step towards the goal of developing professional forecasting models to address similar problems.

The approach used here is the construction of a forecasting model. Forecasts are useful operational inputs for specific government body planning for the years to come (Sohst et al, 2020). The forecasting approach was chosen for three reasons: the first concerns the expertise of Department B in mathematical modelling and the relevance with our research goal, the second concerns the fact that data-driven approaches tend to be explicit in how assumptions within a model can affect future migration flows (Sardoschau, 2020) and that they give a concrete numerical estimation of future flows. The third reason concerns the fact that forecasts are less limited in the time frame they can cover since, in principle, past data trends can be used for many years to come (Sardoschau, 2020). The last reason stems out from the fact that despite Greece has been featuring in some global migration forecasts, nevertheless there is not a concrete prediction model for the unique case of Greece which serves as a forefront country in terms of migration and refugee flows and is subjected to a number of imponderable flow factors due, among others, to its neighborhood with Turkey. More accurately, the few large providers of population projections that include projections of international migration (Migration Data Portal, 2021 October) do not provide sufficient data for Greece or do not provide any data at all: The United Nations Population Division of the Department of Economic and Social Affairs, which covers 233 countries, making it the geographically most complete dataset does not feature Greece in its forecasts. The Wittgenstein Centre for Demography and Global Human Capital, which provides projections of net migration rates per country until the year 2100, provides some relevant data for Greece that do not suffice for drawing management policy scenarios. The US Census Bureau produces net international migration estimates until the year 2060. Again, Greece as a separate European country does not feature in these forecasts. Other datasets are available for particular regions, such as the EUROPOP, produced by Eurostat. These datasets provide information at national level for all European countries. Greece is included in these datasets, but, yet, not all aspects included in the proposed project are taken into consideration such as the issue of imponderable flow factors. From the above it becomes evident that at the present time there are insufficient data on the creation of predictive models of migration flows for the Greek case. Apart from the analysis of legal, political, social and individual evaluation of statistical data, there is no study that proposes a model for forecasting refugee and migration flows in Greece. Hence, this is the challenge this project aspires to tackle. Social sciences have recently been exploiting statistical and machine learning methods to study problems that include, among other tasks, policy making. Although these models cannot replace the scientists, they assist them by handling available data and detecting hidden trends and motifs. Chatsiou & Mikhaylov, present the initial uses of machine learning in social sciences (Chatsiou & Mikhaylov, 2020), and Grimmer et al (2021) in a recent review paper reveal the innovations and the increasing trend of using machine learning techniques on problems of social sciences and the need for synergy between social scientists, mathematicians and computer scientists for solving critical issues in the relative fields. Moreover, departments of social sciences at some precious universities have included statistics and machine learning courses in their syllabus. A characteristic example is a course offered by the department of political sciences of MIT (course 17.835. Details in references). In the proposed project, we shall describe the available data using Hidden Markov Models (HMMs), which belong to the extended area of Machine Learning. HMMs are powerful probabilistic models determined through a few parameters, a property that is essential for efficient performance when we do not have extended data. HMMs have already been successfully used in many fields such as bioinformatics (Eddy, 1998), economics (Yu & Sheblé, 2006), earth sciences (Smaragdakis & Taroudakis, 2020), speech recognition (Srinivasan, 2011), even in social sciences such as psychology (de Haan-Rietdijk et al, 2017).

References

Arsenijević, J., Schillberg, E., Ponthieu, A., Malvisi, L., Ahmed, W. A. E., Argenziano, S., … & de Vingne, B. (2017) A crisis of protection and safe passage: violence experienced by migrants/refugees travelling along the Western Balkan corridor to Northern Europe. Conflict and health, 11(1), 1-9.

Chatsiou, K., & Mikhaylov, S. J. (2020). Deep Learning for Political Science. arXiv preprint arXiv:2005.06540.

de Haan-Rietdijk, S., Kuppens, P., Bergeman, C. S., Sheeber, L. B., Allen, N. B., & Hamaker, E. L. (2017). On the use of mixed Markov models for intensive longitudinal data. Multivariate behavioral research, 52(6), 747-767.

Dimari, G., Papadakis, N. (2022 under publication). Refugee Crisis and Transformations in Greek Migration Policy: the Trend towards Securitization and its Relationship to Precarity, in the case of Greece. Migration, Mobility & Displacement.

Disney, G., Winiowski, A., Forster, J. J., Smith, P. W. F., & Bijak, J. (2015). Evaluation of existing migration forecasting methods and models. (Re-port for the Migration Advisory Committee: commissioned research). ESRC Centre for Population Change, University of Southampton. https://www.gov.uk/government/publications/evaluationof-existing-migration-forecasting-methods-and-models

Eddy, S. R. (1998). Profile hidden Markov models. Bioinformatics (Oxford, England), 14(9), 755-763.

Grimmer, J., Roberts, M. E., & Stewart, B. M. (2021). Machine Learning for Social Science: An Agnostic Approach. Annual Review of Political Science, 24, 395-419.

Guiraudon, V. (2017). The constitution of a European immigration policy domain: a political sociology approach. In Immigration (pp. 285-304). Routledge.

Ministry of Migration and Asylum (2021a March). March Newsletter – Less than 60,000 asylum seekers and refugees now live in Greece. The islands are home to 13,495 out of 40,300 who lived in March 2020. Available at: https://migration.gov.gr/enimerotiko-martioy-2021/ [Accessed 8/5/2021]. (In Greek).

MIT Course 17.835: Machine Learning and Data Science in Politics, https://www.guillermotoral.com/syllabi/machine_learning_syllabus.pdf

Papadakis, N. (2021 forthcoming) Cultural Diversity in the EU: theoretical insights and critical notes on the EU migration policy, in the context of the refugee crisis. In D. Anagnostopoulou (ed), The role of Intercultural Dialogue in managing diversity in Europe. Conference Proceedings. Springer Publishers (under publication).

Sardoschau, S. (2020). The Future of Migration to Germany: Assessing Methods in Migration Forecasting. DeZIM-Institut.

Smaragdakis, C., & Taroudakis, M. I. (2020). Acoustic signal characterization based on hidden Markov models with applications to geoacoustic inversions. The Journal of the Acoustical Society of America, 148(4), 2337-2350.

Sohst, R., Tjaden, J., de Valk, H., & Melde, S. (2020). The future of migration to Europe: A systematic review of the literature on migration scenarios and forecasts. International Organization for Migration: Geneva, Germany.

Srinivasan, A. (2011). Speech recognition using Hidden Markov model. Applied Mathematical Sciences, 5(79), 3943-3948.

UNHCR (2020) Regional Bureau for Europe – Greece Flash Update – Moria Fire Emergency. Available at: https://data2.unhcr.org/en/documents/details/79198 [Accessed 13/5/2021].

Yu, W., & Sheblé, G. B. (2006). Modeling electricity markets with hidden Markov model. Electric power systems research, 76(6-7), 445-451.

Scientific impact: As shown in the literature review, the academic literature in the past few years has given new impulses on how to improve migration forecasting, which, nevertheless, is not yet implemented in most official forecasts. We have also seen why more detailed data could be of great help only to gain a better understanding of the impact of international migration, but also to derive more sophisticated and more accurate forecasts based on these results. Machine learning could be a very interesting polyparametric tool in this perspective. Yet, absent. Yet, such solutions are absent in the Greek case. Therefore, this project will contribute significantly in the scientific ergographies of political sciences and mathematics, as it will generate new interdisciplinary tools and data that could give impetus to a more country specific solution on migration management policy making. These tools consist of the construction of an experimental start-up pilot migration prediction flows model for the Greek case, that could be further developed in the future and the subsequent policy evidence based policy scenario making.

Societal Impact: Forecasting migration is an important requisite for population forecasts or for planning in fields that depend on the future size and structure of the population, such as economics, epidemiology, social insurance, or infrastructure. For the Greece case, this is a very crucial aspect as forecasts will lead to policy scenarios that in turn can lead to the Greek needs assessments in two major fields that concern migration: security aspects and integration ones. Form the one side, the societal impact concerns the host society and its overall security concerns, as forecasts could prove indispensable in the migration management policy making in terms of ‘arrival’ politics and subsequent management of flows according to coordinated polity actions. From the other side, the societal impact concerns the possibility to make management strategic planning pertaining to integration and inclusion of migrants and refugees, by focusing on the translation of the data obtained into indexes of economic, epidemiological, social insurance and infrastructure policy making.

Measures to maximise impact: The aim of the dissemination and exploitation of the project is to create different materials in order to disseminate its findings and to inform both the public, the scientific community and policy makers-specialists on the implementation and findings of the proposed project. The research team will be responsible for coordinating all activities which involve dissemination of project related information to any audience. They will ensure that the results are swiftly disseminated to a large number of relevant academic, public and political organizations.

Target end-users: Primary target end-users are located in Greece mainly and the EU overall (secondary) and they are: Government departments for migration policies, universities, NGOs and research centres. Secondary target users are EU institutions that deal with the issue of migration, such as the European Commission (EC), the European Parliament (EP), the International Organization for Migration (IOM), the International Centre for Migration Policy Development (ICMPD), The Euro-Mediterranean Consortium for Applied Research on International Migration (CARIM), the European Migration Network (EMN), the Fundamental Rights Agency (FRA), the General Director’s Migration Services Conference (GDISC), the Network Health for Undocumented Migrants and Asylum Seekers (HUMA Network), The Migration, Asylum, Refugees Regional Initiative (MARRI), Migration Policy Institute in Europe (MPI Europe). The final target is a mix of all the previous, geographically focused on Greece mainly and in Europe overall. The overall outcomes of the project will be disseminated through the conduction of an international virtual workshop, the participation in one (at least) international conference, one (at least) scientific publication, and the communication material, in order to foster the visibility of the research outcomes and the proposed strategy. The results of the proposed action definitely aim to have a long-term impact in the sense that the start-up pilot model will be further expandable. The strategy and the outcomes will be disseminated largely through all the channels mentioned above and will be also provided to policy makers in order to evaluate them and implement the necessary interventions. For these reasons the impact of this action will be direct and also long-term in the area under study. Communication activities will be focused on divulging the project itself, acknowledging the support of University of Crete in fostering and funding interdisciplinary research initiatives.

The model will use data that quantify the migration flows in combination with data concerning the adaptation of the immigrants to their host nation and the specific circumstance that caused the necessity for them leaving their home country. Here we will restrict our analysis to Greece as the host country, but immigrants will be of any nation represented in the immigrant population.

The selected model would be a multivariate predictive model. Such models exploit multidimensional data (multidimensional time-series) to predict future values of a variable (usual univariate or bivariate time-series) under consideration. The target variable will be the number of immigrants flows (in, out) in specific time horizons in our study. The parameters that will describe the migration flows are demographics, migration paths, political conditions. Furthermore, we will collect and quantify information about the political conditions and natural disasters that occurred in the involved nations.

There are many methodologies for challenging such prediction tasks. When the data are adequately large, a Recurrent Neural Network (RNN) is a robust choice. However, the lack of detailed multi-parametric data, as might be the case in our project, makes such a model not a good candidate for our application.

According to the size of the available data, a strong candidate for the task at hand is Hidden Markov Models (HMM). These probabilistic models provide good performance at modelling the stochastic behaviour of complex systems by introducing hidden variables that are considered responsible for producing the observations. The concept is to consider hidden variables with three hidden states; each represents a crisis scenario (soft, medium, hard), and the model would generate a prediction for a specific time according to the emission distribution corresponding to the hidden state. The type of the emission distribution that we will use to the HMM has to be investigated. HMMs have been used by Smaragdakis and Taroudakis for the classification of underwater and seismic signals  (Smaragdakis and Taroudakis, 2016, 2020a, 2020b). The expertise thus gained will be exploited in the problem under consideration, using the basic concepts of the already applied techniques. The main challenging of the adaptation of this technique to problems from Social Sciences is the appropriate and scientifically correct association of the data with the physical parameters determining the trends of the event under study.

We will use a Monte-Carlo approach to sample future values of the hidden variables of the probabilistic model for single-step and multi-step predictions. More specifically, as we anticipate making predictions corresponding to the different crisis scenarios, we will perform three different batches of simulated realities by imposing to the very next time step hidden variables all the available transition patterns and predicting the ones of the next time steps up to the considered time horizon. Then, for each simulated step, we will get a sample of the variables of interest (e.g. number of incoming immigrants at a specific time step) from the emission distributions.

As HMMs can only describe stationary time-series, we will investigate if the data present monotone trends. In such  scenarios, we will detect and will model the trends using proper mathematical functions. Furthermore, scaling and standardizing data is the norm in such models. The final predictions could then occur by combining the predictions of the HMM and the ones of the modelled trends.

The validation of the model will be performed by considering the immigration flows in Greece since 1990 or so. First, we will use the data up to 2019 in the training phase of the model and then run predictions up to the date of running. Once we have ensured an adequate performance of the model, we will retrain it using the whole dataset to provide its full potential to predict future immigration flows in the Greek region. The model could be re-adapted on-line and constantly provide the interested researcher with up-to-date forecasts.

We should notice that the entire project will serve as the preliminary research step toward the ambitious task of tackling the complex problem of forecasting immigration flows in an area of interest, and therefore the resulted model should be considered a pilot one.

The implementation of the model will be held by a computer program that will exploit multiple-CPU cores. The program will be developed using the Python and Rust programming languages and the TensorFlow machine learning library.

References

Smaragdakis C. and Taroudakis M. (2016): “Hidden Markov Models feature extraction for inverting underwater acoustic signals using wavelet packet coefficients” in Proceedings Euroregio 2016, Porto.

Smaragdakis C. and Taroudakis M.I. (2020): «Probabilistic Characterization of Acoustic ad Sesimic Signals». Ercim News, Vol 122, pp 35-36.

Smaragdakis C. and Taroudakis M.I. (2020): “Acoustic signal characterization based on hidden Markov Models with applications to geoacoustic inversions»  Journal of the Acoustical Society of America Vol. 148, pp 2337-2350.

Month

Month

Description of Work: The outputs include a meeting where the action plan will be designed. The overall man-agement will be conducted during the period of the implementation of the project through meetings and evalu-ation and direct coordination from the project manager (Coordinator). There will be also an Interim Evaluation in order to indicate the problems and best practices and a final evaluation meeting that will conclude to the final report of the project. The management of the project will be conducted by the Coordinator and the Principal Investigator of the project, along with their two supervisors. Monitoring of progress will be carried out through reports, meetings and minutes. The Coordinator will keep detailed records of input into the project using logbooks and a time-sheet system. The Coordinator and the Principal Investigator of the project will set up an internal server system, which can be accessed by both. The server will be password protected, enable access hierarchy and serve as a repository of all documents of the project. Frequent and informal communications be-tween the two will be encouraged at all times – via phone, internet conference calls and e-mail, primarily.

Tasks:

1.1 Interim Report of the project that will evaluate the work done and assess any problems and inefficiencies occurred

1.2 Final Evaluation Report. At the end of the project based on the findings and the total implementation of the project’s actions in order to evaluate the actions and foster the post-project usage of the outcomes as well as any necessary feedback

Deliverables1:

D1. First Report of Progress (M9),

D2. Overall report of findings (M18)

Milestones: 1: Overall report of findings (M18)

Description of Work: This WP is dedicated to the examination of quantitative data from the 1990s when Greece became a host country of migrants up to 2021, when the results of the New Pact on Migration and Asylum have started to become visible in terms of flows and as such a comparative assessment on migrations flows can be made. In addition, a literature review pertaining to the migration policy making the period following the 2015 refugee crisis (five nodes of transformation of Greek Migration Policy) will be conducted in tandem with a litera-ture review that will include data on migration prediction models, their usage and implementation at a Europe-an level. The last phase of the work carried out in this project entails the articulation and justification of the three migration flows crises scenarios in order to construct the prediction model and make policy scenarios. These are the hard flows scenario which pertains to events such as wars and catastrophic events that could trigger massive migration and refugee flows in Greece. The foci here are the Middle East and Africa. The Medium flows scenario pertains to manageable numbers of migration whereby there exists certain ‘normality’ in the arrival of migrants that come in Greece mainly for job reasons. Last, the soft flows scenario pertains to events that trigger the decrease of migrants/refugees flows in Greece, such as for instance, pandemics.

Tasks:

2.1 Review and analysis of quantitative data from the 1990s when Greece became a host country of migrants up to 2021.

2.2 Review and analysis of state of play of migration policy making in Greece for the period 2015-2021 (five nodes of transformation of Greek Migration Policy)

2.2 Review and analysis of current models of prediction of migration flows.

2.3 Articulation of three prediction of migration flows scenarios: Hard scenario, whereby sudden mega-events could trigger massive refugee flows to Greece (events such as the recent turmoil in Afghanistan, the upcoming election of Turkey in 2023, climate change and famine in Africa will be taken into consideration) medium scenar-io, whereby migration flows in Greece are considered manageable (here the foci is on economic migration and not refugee crises) and soft scenario whereby a smooth reduction of migration and refugee flows is observed (one such case is the occurrence of pandemics).

Deliverables:

D3. Synthetic Report that will entail a timeline of migrations flows for the Greek case from the 1990s to 2021, a literature review on the state of play of migration policy making in Greece for the period 2015-2021 (five nodes of transformation of Greek Migration Policy) in addition to data concerning current models of prediction of mi-gration flows their usage and implementation at a European level (M3).

D4: Technical report on three prediction of migration flows scenarios: Hard scenario, whereby sudden mega-events could trigger massive refugee flows to Greece (events such as the recent turmoil in Afghanistan, the up-coming election of Turkey in 2023, climate change and famine in Africa will be taken into consideration as well) medium scenario, whereby migration flows in Greece are considered manageable (here the foci is on economic migration and not refugee crises) and soft scenario whereby a smooth reduction of migration and refugee flows is observed in case of mega-events as well (one such case is the occurrence of pandemics) (M7).

Milestones:

2: Synthetic Report of literature review (M3).

3: Technical Report on three prediction of migration flows scenarios (M7)

Description of Work:

In this WP, we will develop, implement and fine-tune the migration flows prediction model. We will start by collecting and processing the data.

Then we will develop the mathematical framework of the model that includes the determination of the model parameters that control the selected probabilistic model. In particular, as we will use hidden Markov models, we have to pre-decide the number of the hidden states and the emission distributions. Although we will be enforced three hidden states, one per policy scenario, we will test various distributions for the emissions to conclude a choice that better matches the data.

The proper variables will be selected in connection with the WP3 and must be a dynamic process. The political scientist of the team will constantly be evaluating the model’s performance using controlled data by comparing the model’s outputs with the expected outcomes. The final selection of the variables would result from a fine-tuning process until the expected performance of the model to be achieved.

During this WP, we will develop the appropriate computer program for testing and releasing the prediction model.

Tasks:

3.1 Collecting and (pre-)processing the selected data, which will be the input data of the model.

3.2 Developing the mathematical framework for the predicting model.

3.3 Writing the computer program that implements the proposed predicting model.

3.4 Testing the prediction model with controlled data for fine-tuning both the mathematical and the computational aspects of the model.

Deliverables:

D5: Migration Flows Prediction Model including a user’s manual (M15)

Milestones:

4: Migration Flows Prediction Model (M15)

Description of Work: This WP will work as such: at first instance, each set of data obtained from the articulation of the three different crises scenarios (hard, medium, soft) will be inserted separately to the Migration flows Prediction Model. The model will produce for each of the three scenario another set of predictive statistical and numerical data which will then be taken to be developed into a comprehensive report that will lead to the artic-ulation of three different migration scenarios. The first migration policy scenario will reflect the predictions per-taining to the hard crises scenario, the second migration policy scenario will reflect the predictions regarding the medium crises scenario whereas the last migration policy scenario will reflect the predictions of the soft crises scenario.

Tasks:

4.1 Insertion of data of hard flows crisis scenario into migration flows prediction model

4.2 Insertion of data of medium flows crisis scenario into migration flows prediction model

4.3 Insertion of data of soft flows crisis scenario into migration flows prediction model

4.4 Drafting of hard flows report

4.5 Drafting of medium flows report

4.6 Drafting of soft flows report

4.7 First Migration Policy Scenario (Hard)

4.8 Second Migration Policy Scenario (Medium)

4.9 Third Migration Policy Scenario (Soft)

4.10 Synthesis of findings into one overall technical report pertaining to Migration Management Policy Scenarios.

Deliverables:

D6: Technical Report of Hard, Medium, Soft Migration Policy Scenario  (M18)

Milestones: 5: Technical Report titled: “Migration Management Policy Scenarios for three Different Flows Crises Scenarios: Hard, Medium and Soft (M18)

Description of Work: This WP entails a series of dissemination activities that are expected to commence the third month of the implementation of the project up to its closure. These include the preparation of dissemination material such as a logo, brochure and a poster for the project, the preparation of (at least) one article publication in a peer reviewed journal, the conduction of an international virtual workshop and the participation of the research team to at least one international conference.

Tasks:

5.1 Preparation of dissemination material (logo, brochures, poster)

5.2 Journal publication preparation

5.3 Participation in international conference

5.4 Conduction of an international virtual workshop

Deliverables:

D.7 Publication of one (at least) scientific article (M18)

D.8 Participation in international conference (M16)

D.9 Conduction of an international virtual workshop (M18)

Milestones:

5. Publication of one (at least) scientific article (M18)

6. Participation in international conference (M16)

7. Conduction of an international virtual workshop (M18)

Papadakis N. (2021), “Cultural Diversity in the EU: theoretical insights and critical notes on the EU migration policy, in the context of the refugee crisis”, in D. Anagnostopoulou (ed), The role of Intercultural Dialogue in managing diversity in Europe. Conference Proceedings. Springer Publishers (under publication).

Papadakis N., Kosmadakis N. (2021), “Multiculturalism and ethno-cultural diversity: theoretical insights and challenges for the current European Migration Policy”, in D. Kotroyannos & St. Tzagkarakis (eds), Prospects for the Social and Economic Integration of Refugees in Greece. Athens: Centre for Human Rights (KEADIK) of the University of Crete and I. Sideris publ. (under publication).

Dimari, G., Papadakis, N. (2022 under publication). Refugee Crisis and Transformations in Greek Migration Policy: the Trend towards Securitization and its Relationship to Precarity, in the case of Greece. Migration, Mobility & Displacement.

Papadakis N. & Fragkoulis I. (2007), “Moral liberalism and migration policy. International trends and challenges for the Greek case”, in X. Kontiadis & Th. Papatheodorou (eds.), The Reform of Migration Policy. Athens: Papazisis, pp. 145-156.

Dimari, G. (2021). Desecuritizing Migration in Greece: Contesting Securitization through “Flexicuritization”. Interna-tional Migration, 0 (0), 1-15. https://doi.org/10.1111/imig.12837.

Dimari, G. (2020). The Securitization of Migration in Greece 2011-2019: A Discourse and Practice Analysis. European Quarterly of Political Attitudes and Mentalities, 9(4), 1-13.

Smaragdakis C. and Taroudakis M. (2016): “Hidden Markov Models feature extraction for inverting underwater acoustic signals using wavelet packet coefficients” in Proceedings Euroregio 2016, Porto.

Smaragdakis C. and Taroudakis M.I.: (2020) «Probabilistic Characterization of Acoustic ad Sesimic Signals». Ercim News, Vol 122, pp 35-36.

Smaragdakis C. and Taroudakis M.I.: (2020) “Acoustic signal characterization based on hidden Markov Models with applications to geoacoustic inversions»  Journal of the Acoustical Society of America Vol. 148, pp 2337-2350.