The coronavirus pandemic of 2019 and 2020 and the civil rights crisis of 2020, led by the Black Lives Matter movement, have highlighted some of the major limitations of our society today.

在黑人生活问题运动的带动下，2019年和2020年的冠状病毒大流行以及2020年的民权危机凸显了当今社会的一些主要局限性。

While our economy is growing at an increasingly rapid pace, many areas of development remain under-considered. A global vision of the interconnection between the different facets of development would help to achieve a sustainable improvement in living conditions.

尽管我们的经济以越来越快的速度增长，但许多发展领域仍未得到充分考虑。 对发展的不同方面之间相互联系的全球视野将有助于实现生活条件的可持续改善。

“[Sustainable Development is:] Development that meets the needs of the present while safeguarding Earth’s life-support system, on which the welfare of current and future generations depends.” [1]

“ [可持续发展是：]在满足当前需求的同时保护地球生命维持系统的发展，这是今世后代的福祉所依赖的。” [1]

17个可持续发展目标 (The 17 Sustainable Development Goals)

The SDGs cover seventeen interconnected goals that define global quantifiable objectives across the social, economic, and environmental dimensions of development. [1] They aim to be a universal set of indicators and a reference framework to be leveraged by the global community to motivate policies and implementation by 2030.

可持续发展目标涵盖了十七个相互关联的目标，这些目标定义了发展的社会 ， 经济和环境方面的全球可量化目标。 [1]它们旨在成为一套通用的指标和参考框架，国际社会将利用它们来在2030年之前激励政策和实施。

They were set by the United Nations in 2015 as a follow-up to the Millennium Development Goals, which provided a framework for a global partnership to reduce extreme poverty.

它们是联合国在2015年设定的，是千年发展目标的后续行动，该目标为减少极端贫困的全球伙伴关系提供了框架。

This new framework supports a long term transition towards more sustainable development. It fosters accountability while also promoting global collaboration. It is a tool to guide decision making but not in and of itself a prescriptive and actionable guide. As a result, member states and institutions are free to pursue policies and programs according to their own context, resources, and available scientific evidence.

这个新框架支持向更可持续发展的长期过渡。 它在促进问责制的同时还促进了全球合作 。 它是指导决策的工具，但它本身并不是规定性和可操作性的指南。 结果，成员国和机构可以根据自己的背景，资源和可用的科学证据自由地制定政策和计划。

互连与责任 (Interconnections and accountability)

The global goals highlight how the 17 goals are interconnected. The increase in poverty is likely to lead to a decrease in hygiene and an increase in health challenges. Meanwhile, with poverty, access to education is harder and violence may emerge more easily. Similarly, Climate Change endangers wildlife (on land and below water) and will increase the already existing inequalities between people. This explains why Griggs D, Staord-Smith M, Ganey O, et al. [1] suggested that special emphasis should be placed on earth life support systems and poverty reduction.

全球目标突出了这17个目标之间的相互联系 。 贫困的增加可能导致卫生状况的下降和健康挑战的增加。 同时，由于贫困，受教育的机会更加困难，暴力事件也更容易出现。 同样，气候变化危害野生动植物(陆地和水下)，并会加剧人与人之间已经存在的不平等现象。 这解释了为什么Griggs D，Staord-Smith M，Ganey O等人。 [1]建议应特别重视地球生命支持系统和减贫 。

“For the economy itself will die if the ecosystems collapse” Leonardo Di Caprio, UN Messenger of Peace

“如果生态系统崩溃，经济本身将死亡”，联合国和平信使莱昂纳多·迪卡普里奥

Sustainability areas can be considered as a large organism whose parts are interconnected. The economy exists within society, which in turn exists within the global ecosystem of the Planet Earth. This framework highlights that the economy cannot exist without society, which cannot function without the environment that surrounds it.

可持续发展领域可以看作是相互联系的大型生物 。 经济存在于社会内部，而社会又存在于地球的全球生态系统中。 该框架强调，没有社会就不可能存在经济，没有社会就不能运转经济。

从愿景到实际行动和持久变化 (From the vision to actual actions and lasting changes)

Global goals have been defined to inspire leaders. Yet, 17 general goals with 169 targets are not in itself sufficient to trigger lasting impactful decisions. In order to move towards more sustainable countries and businesses, we need to be aware of what changes will have a lasting impact. This requires global partnerships (Goal 17) as well as centralization and easy access to information.

制定了全球目标以激励领导者。 但是，17个总体目标和169个目标本身不足以触发持久的有影响力的决策 。 为了朝着更具可持续性的国家和企业迈进，我们需要意识到哪些变化将产生持久的影响。 这需要全球伙伴关系 (目标17)以及集中化和方便地获取信息。

From the community of researchers, scientists, entrepreneurs, and innovators, a large amount of knowledge, ideas, and new products are created on a daily basis, which helps in part to achieve the SDGs. However, much work needs to be done to gather knowledge and information and to make it usable by stakeholders and decision-makers.

在研究人员，科学家，企业家和创新者的社区中，每天都会产生大量的知识，思想和新产品，这在一定程度上有助于实现可持续发展目标。 但是，需要做很多工作来收集知识和信息，并使其可由利益相关者和决策者使用。

For example, if a decision-maker is wondering: what can I do to improve education while reducing inequalities in the population? A clear set of scientific and business evidence should be readily available to support brainstorming on the solution.

例如，如果决策者想知道： 在减少人口不平等的同时，我该怎么做才能改善教育水平？ 一组清晰的科学和商业证据应随时可用，以支持就解决方案进行集思广益。

Therefore, some additional information would help stakeholders to adopt sustainable changes:

因此，一些其他信息将有助于利益相关者采取可持续的变革：

• good and centralized access to historical data relevant to the goals

  良好且集中地访问与目标相关的历史数据

• scenarios and guidance on the impact of different policies or business changes, with evidence of positive and negative impact on different goals

  有关不同政策或业务变更影响的方案和指南 ，并证明对不同目标有正面和负面影响

• centralizing and summarizing researches and patents that are relevant to the goals

  集中和总结与目标相关的研究和专利

With particular emphasis on the interconnection and mutual impact of the different scenarios.

特别强调不同方案的相互联系和相互影响。

利用数据和人工智能实现可持续发展目标 (Leveraging Data and Artificial Intelligence for the SDGs)

Artificial Intelligence will play a major role in improving access to knowledge. In particular, text mining and natural language processing will help to make sense of the ever-increasing amount of information published daily by the research community, industry, and governments.

人工智能将在改善知识获取方面发挥重要作用。 尤其是， 文本挖掘和自然语言处理将有助于弄清研究界，行业和政府每天发布的信息量的不断增长。

趋势扫描仪 (The Trendscanner)

One such initiative, led by the United Nations Development Programme (UNDP), the RISE Research Institutes of Sweden, and the AI for Good Foundation, is Trendscanner. This visualization browser maps major current trends such as IoT, Urbanization, and Big Data to the SDGs and brings together relevant reports and documents.

Trendscanner是由联合国开发计划署(UNDP)，瑞典RISE研究机构和AI for Good Foundation领导的一项此类倡议。 该可视化浏览器将IoT，城市化和大数据等当前的主要趋势映射到SDG，并将相关报告和文档汇总在一起。

SDG数据目录 (The SDG Data Catalog)

With this in mind, together with James H. and the AI for Good Foundation, we are building the Sustainable Development Data Catalog, an open, extensible, global database of data sets, metadata, and research networks built automatically by mining millions of published open access academic works. The SDG Data Catalog is focusing on the issue of data discovery and data sharing to accelerate progress on the Sustainable Development Goals.

考虑到这一点，我们将与James H.和AI for Good Foundation一起，构建可持续发展数据目录 ，这是一个开放的，可扩展的，全球性的数据集，元数据和研究网络数据库，通过挖掘数百万已发布的公开内容而自动建立访问学术作品。 可持续发展目标数据目录的重点是数据发现和数据共享问题，以加快实现可持续发展目标的进度 。

The process to generates this catalog is summarized in the system design below. From open-access databases gathering research papers information, a large dataset of published papers is created (5 Million papers in our first version). Those papers are then parsed to extract datasets mentions and relevant metadata associated with it.

下面的系统设计中总结了生成此目录的过程。 从收集研究论文信息的开放获取数据库中，可以创建大量已发表论文的数据集(我们的第一个版本中有500万篇论文)。 然后对这些文件进行解析，以提取提及的数据集和与之相关的相关元数据。

The extraction phase is using a flexible web-scraper that is both scalable, light-weight, and copyright aware. The process has a 65% success in PDF retrieval.

提取阶段使用的是灵活的网络抓取工具，该工具具有可伸缩性，轻巧性和版权意识。 该过程在PDF检索中取得了65％的成功。

The main component of the information extraction pipeline is the named entity recognition model that identifies dataset names, dataset description, owner, number of samples, and other references and document metadata.

信息提取管道的主要组件是命名实体识别模型 ，该模型识别数据集名称，数据集描述，所有者，样本数量以及其他参考和文档元数据。

Both the classifiers and NER models leverage an active learning strategy [3]. They start with a few manual annotations that are gradually exploited to train the model, suggest examples for annotation where the uncertainty is highest, annotate some more instances, and improve the overall accuracy of the task. This is made possible thanks to the Prodigy software [4].

分类器和NER模型都利用主动学习策略 [3]。 他们从一些手动注释开始，这些注释逐渐被用来训练模型，为不确定性最高的注释示例提供注释，对更多实例进行注释，并提高了任务的整体准确性。 这要归功于Prodigy软件[4]。

As the next steps, we wish to build query and visualization layers on top of the SDG Data Catalog. In the end, we hope to create an easily navigable Wikipedia-like experience for keeping track of useful data.

下一步，我们希望在SDG数据目录的顶部构建查询和可视化层。 最后，我们希望创建一种易于导航的类似Wikipedia的体验，以跟踪有用的数据 。

数据公平化 (FAIRification of data)

While the amount of scientific data created is enormous, many research publications do not contain detailed descriptions of the data sets used and how to access them. This has multiple explanations. First, some data sets have great scientific values but will not give credit as such, so scientists may prefer to keep them private in order to exploit their analytical potential. [5] In addition, many data sets are privately held and retained as a competitive advantage by commercial entities.

尽管创建的科学数据量巨大，但许多研究出版物并未包含对所使用的数据集以及如何访问它们的详细描述。 这有多种解释。 首先，某些数据集具有很高的科学价值，但不会因此而受到赞誉，因此科学家们可能更愿意对其保密，以发挥其分析潜力。 [5]此外， 许多数据集由商业实体私有拥有并保留为竞争优势。

The FAIR Data Project’s Commitment Statement in the Earth, Space, and Environmental Sciences [6] has been created to address this problem. FAIR data stands for Findable, Accessible, Interoperable, and Reusable. This commitment makes it more difficult to be published in most journals in these fields without including datasets that are used in a way that supports the FAIR principles.

FAIR数据项目在《地球，空间与环境科学》中的承诺声明[6]已创建，用于解决此问题。 FAIR数据代表可查找 ， 可访问 ，可互操作和可重用 。 这种承诺使得在不包含支持FAIR原则的方式使用的数据集的情况下，很难在这些领域的大多数期刊中发表。

Making data FAIR will improve research and science in several ways. It is a move towards open-source and the availability of the most recent discoveries and will make most science publications more reproducible and comparable.

使数据公平将以多种方式改善研究和科学。 这是向开源和最新发现的可用性迈进的一步，它将使大多数科学出版物更具可复制性和可比性 。

Our work to build the SDG data catalog contributes to making SDGs data FAIR.

我们构建SDG数据目录的工作有助于使SDG数据公平 。

结论 (Conclusion)

If not an actionable guide toward more sustainability, at least the global goals are a common language that companies, NGOs, and governments around the world can use to frame their policies and decisions. As Scientists, Researchers, Innovators, and Entrepreneurs, it is our role to make the bridge between the goals and the decisions that will lead to the necessary changes.

如果不是切实可行的指导以实现更多的可持续性，至少全球目标是世界各地的公司，非政府组织和政府可以用来制定其政策和决策的通用语言 。 作为科学家，研究人员，创新者和企业家，在目标和决定之间架起桥梁，这将导致必要的改变，这是我们的职责。