December 2020，Information Polity, the Dutch journal, launched a special issue "Algorithm transparency in the government: Algorithmic Transparency in Government: Towards A Multi-Level Perspective).。Six of them analyzed the challenges experienced by different levels of governments in the process of using algorithms for decision -making，It aims to help people better understand how to use algorithms in government decisions at the concept and empirical level。

　Bet365 app download

According to special issues of special issues、Assistant professor of Public Management College of the University of Public Management of the Netherlands Sarah Gieste and the deputy professor of the School of Governance of the University of Undelch University, Stephan Grimmelikhuijsen, introduced，Research in the past，Machine learning technology and algorithms are expected to improve the fairness of government services、Effective，"Relieve" decision from human subjectivity。Now，The algorithm is medical、Judicial and other public service areas are becoming increasingly popular。For example，In the judicial system，The accuracy of the probability of the criminal re -of -crime by the algorithm predicts that the probability of the criminal is higher than the human judge。

but，Agent -based decision Bet365 lotto review -making or suggestions may also bring too many negative effects on the disadvantaged group。For example，Some local governments in the Netherlands have introduced a system，Filtering people with algorithms that may implement social welfare fraud。The application of this system has been opposed by some people，and was determined by a court in the Netherlands in 2020, "discriminatory" and "opaque"，Because disadvantaged groups are often predicted as actors suspected of fraud。A major problem behind this phenomenon of bias (BIAS) is the lack of transparency of the algorithm。Actually，The algorithms used in government decision -making are mostly developed by commercial institutions，These algorithms are considered intellectual property rights，It will not be disclosed to the people。The government's decision -making based on opaque algorithms is risky，Because of biased decisions, it will erode the people's trust in the government。Gust talk about，Critical criticisms for the decision -making of government algorithms rapidly increase，It is not enough to reveal the technical details of the algorithm，​​It is also necessary to examine the system operating algorithm、Institution、Personal background，This can really ensure that the government is transparent、Use algorithms responsibly。

　Transparency involves various government levels

Gramrick Huaison said，The transparent challenges used by algorithm running through the different levels of the government。Legislative agencies can bet365 best casino games set up algorithm transparency mechanisms，Administrative agencies can introduce new policies used by algorithms，Government staff can have their own knowledge and skills library，All these levels interaction and affect algorithm transparency in the public institution。Transparency challenges may also present different forms，Technical tools、Regulatory guidance policy、Institutional policy can increase transparency，can also limit transparency。Articles included in the special issue are in macro (macro-)、Middle View (Meso-)、Micro (Micro-) three levels of three levels discussed transparent challenges。Macro -level description Phenomenon under the perspective of the system: National system、Laws and regulations、What role does culture play in algorithm decisions? Intermediate -level followers and teams: How does government agencies generally deal with the transparency of algorithm?、motivation、Interactive、Behavior and other personal attributes，For example，How does community staff use the algorithm? Will the transparent algorithm make them make different or better decisions?

At the macro level，Persti Ahonen, a professor of political science at the University of Helsinki, Finland and Tero Erkkil (Tero Erkkil?)。The official documents of national and international documents are analyzed，Carry out semi -structured expert interviews，They found that the laws and regulations of the European Union and Finland intertwined in a complex way。Although Bet365 lotto review there are binding specifications，For example, "General Data Protection Regulations" (GDPR)，But how member states have a certain free space in how to formulate and implement specific policies。Domestic in Finland，The transparency concept of algorithm transparency and other automated decisions is in the law、Ethics、Politics、Policy、Management、Complex junction zone in the technical field，People have conflict with the interpretation of relevant laws。A Hohonin and Ejela's research provides two important revelations: first，On the transparency problem of algorithm and algorithm decision -making，EU and national legal frameworks and the public domains of member states (PUBLIC SPHERE) traditions in the tradition of both restrictions and promotion factors，Should carry out comparison research on this theme。Second，In order to allow researchers to have greater supervision and control strength，Provide professional opinions at the same time，It is necessary to examine the role of an expert consulting agency。

Assistant professor of the School of Public Management of the University of Public Management of the Netherlands Alex Ingrams (Alex Ingrams) uses UNSUPERVISED MACHINE Learning) to analyze the number received by the US Transportation Administration (TSA) Thousands of public comments。Analysis，Researchers have discovered the top topics (Topic Cluster)，Can help policy makers know a lot of information in the process of public solicitation。This study indicates，Not only the algorithm itself should be transparent，The algorithm can also be used as a Bet365 lotto review tool for improving the transparency of government decision -making。

At the level of the middle view，Qualitative research methods based on semi -structured interviews and literature analysis，Professor of Political Science and International Relations at Madrid Autonomous University in Spain J. Ignacio Criado and other scholars analyzed an algorithm system used by the Valencia Autonomous Region Government。Do they want to answer two questions: To what extent the algorithm affects the freedom of discretion when government staff make decisions? How does the transparency of algorithms affect the freedom of government staff?，The algorithm has a positive effect on the free discretion of government staff。Let them actively participate in the research and development or review process recognized by the law，helps to enhance algorithm transparency。

Frank Bannister, associate professor of the information system of the Saint Trinity University of Ireland, and Regina Connolly, a professor of information system at the Urban University information system at the Irish City University,，With the development of artificial intelligence，Human decision -making is increasingly relying on algorithm and machine learning technology。Risks that can detect algorithm decisions and classify risks，For systematic ground、Coping risk accurately is important。

At the micro level，Mexican Economic Research and Education Center (CIDE) Public Management professor Rik Peeters proposed，If the algorithm can not understand how human decision bet365 live casino games -making，How to design algorithms affects transparency and the actual possibility of human independent judgment，"Keep Humans in the Loop" appeal may be weak。By reviewing the recent academic literature，Pitz discovered，The prerequisite for transparency and self -judgment of human beings by three key design variables (degree of automation、Algorithm type、Organization scope) Decisions，Bounded Rationality、Satisficing Behavior、Automation Bias、Frontline Coping Mechanism on how to use the supervision and Override option in humans on algorithm。

Scholars such as Justin Bullock, deputy professor of Public Service and Management Department of Texas University of Agricultural Engineering, designed an artificial intelligence about artificial intelligence、Human autonomous judgment、The theoretical framework of the research of the administrative system of public institutions integrated，also inspected two very different areas of the public sector using artificial intelligence (medical insurance audit and police)。​​Research results show，The impact of artificial intelligence on human independent judgment is non -linear、non -monotonous (NONMONOTONIC)。

Gest and Gramrick Huaishan said，For the responsibility of advocating、Transparent algorithm use，The next study should explore the above macro、Middle view、Interaction between micro levels。