Data scientists regularly use APIs (Application Programming Interfaces) to implement advanced functionalities without having to build things from the ground up, import data from external sources, expose data for users to consume, or achieve other objectives.

数据科学家经常使用API​​(应用程序编程接口)来实现高级功能，而无需从头开始构建事物，从外部源导入数据， 将数据公开给用户使用或实现其他目标。

For example, if you intend to create your own classifier that categorizes the genre of a piece of music, you can use the Spotify API to do that.

例如，如果您打算创建自己的分类器来对音乐的流派进行分类，则可以使用Spotify API来实现。

You’ll be able to tap into the massive data that Spotify has and build the classifier fast, without having to start from scratch. In such cases, APIs offer the right solution to your data science project.

您将能够利用Spotify拥有的海量数据并快速构建分类器，而无需从头开始。 在这种情况下，API为您的数据科学项目提供了正确的解决方案。

And, to make the most of APIs, there is a need to enforce API management. With an API management platform, such as the Rakuten RapidAPI Enterprise Hub, you can abstract most of the day-to-day administrative pain points of creating or consuming APIs and realize increased productivity.

并且，为了充分利用API，需要实施API管理。 借助Rakuten RapidAPI Enterprise Hub等API管理平台，您可以抽象出创建或使用API​​的大部分日常管理难题 ，并提高生产率。

These solutions assist you in designing, implementing, monitoring, securing, analyzing, managing, and growing your API program.

这些解决方案可帮助您设计，实施，监视，保护，分析，管理和扩展API程序。

In this article, I’m going to talk about how an API management platform can help in enhancing your data science projects.

在本文中，我将讨论API管理平台如何帮助您增强数据科学项目。

1.提供集中的可见性 (1. Provides centralized visibility)

An API management tool will provide you with centralized visibility into the performance of your internal APIs and external API subscriptions. As you use more APIs in your data science projects, it becomes difficult to keep track of them.

API管理工具将使您可以集中查看内部API和外部API订阅的性能。 当您在数据科学项目中使用更多API时，很难跟踪它们。

In fact, according to a 2017 study, an average company uses more than 300 different APIs. With these many APIs to take care of, simplifying their tracking process is required to optimize their value and deliverability.

实际上，根据2017年的一项研究 ，一个普通的公司使用300多种不同的API。 有了这些许多API，就需要简化其跟踪过程以优化其价值和可交付性。

Otherwise, it can be difficult to assess and improve their performance, enhance the developer experience, or make informed decisions.

否则，可能难以评估和改善其性能，增强开发人员的经验或做出明智的决定。

An API management solution will give you the enhanced visibility and control you need to know what is happening with your APIs.

API管理解决方案将为您提供增强的可见性和控制力，您需要了解API所发生的情况。

With a single central location to manage your data science APIs, you can get rid of redundant APIs, identify performance inconsistencies, and ease the orchestration of the different moving parts of your APIs.

通过一个中心位置来管理数据科学API，您可以摆脱冗余的API，识别性能不一致的地方，并简化API不同移动部分的编排。

2.改善API治理 (2. Improves API governance)

API governance refers to the practice of establishing various policies that control the use of an API product.

API治理是指建立控制API产品使用的各种策略的实践。

It involves a wide range of management, administration, visibility, monitoring, and other measures that streamline and govern the adoption of APIs.

它涉及广泛的管理，管理，可见性，监视和其他简化和控制API采用的措施。

As a data scientist, having an API management platform will let you outline the terms and conditions that make your APIs accessible by different personas within and outside your enterprise.

作为数据科学家，拥有API管理平台可让您概述使企业内部和外部的不同角色可以访问API的条款和条件。

It’s what you need to implement access management and safeguard against unauthorized use, undertake audit trails to discover abnormal usage trends, and blacklist or whitelist APIs.

这是实现访问管理，防止未经授权使用，进行审计跟踪以发现异常使用趋势以及黑名单或白名单API所需要的。

Furthermore, managing your APIs makes it easy to analyze and monitor their performance.

此外，通过管理API，可以轻松分析和监视其性能。

Most of these solutions come with intuitive dashboards that allow you to get quick insights across all APIs used by your team. This way, you can evaluate and maintain the health of your APIs.

这些解决方案中的大多数都带有直观的仪表板，可让您快速了解团队使用的所有API。 这样，您可以评估和维护API的运行状况。

You can also use such a platform to collate APIs according to their different functions. For example, if you have different APIs that pull data sets for data visualization, you can group them and monitor their performance easily.

您还可以使用这样的平台根据API的不同功能来整理它们。 例如，如果您有不同的API提取数据集以进行数据可视化 ，则可以将它们分组并轻松监视其性能。

Also, even if there are multiple versions of an API, such a platform will allow you to track and govern the various versions effortlessly.

此外，即使API有多个版本，该平台也可以使您轻松跟踪和管理各种版本。

3.增强API发现 (3. Enhances API discovery)

Most API management platforms come with a built-in developer portal that allows users to find APIs, learn how to consume them, test them out interactively, and ultimately register to get API keys.

大多数API管理平台都带有一个内置的开发人员门户，该门户允许用户查找API，学习如何使用它们，进行交互测试以及最终注册以获得API密钥。

These portals act as a repository of everything pertaining to using the given APIs — from providing documentation, code snippets, tools, and other helper resources.

这些门户充当与使用给定API有关的所有内容的存储库-从提供文档 ，代码段，工具和其他帮助程序资源开始。

If you can use the portal to search for public and internal APIs, sign up to them, and incorporate them into your applications, you can greatly improve the capabilities of your data science projects.

如果您可以使用门户网站搜索公共API和内部API，对其进行注册并将它们合并到您的应用程序中，则可以极大地提高数据科学项目的功能。

Instead of having to create functionalities from scratch, you can use an API management solution to discover new APIs and save your development time, costs, and efforts. This will allow you to concentrate on improving the offerings of your core product.

无需从头开始创建功能，您可以使用API​​管理解决方案来发现新的API，并节省开发时间，成本和工作量。 这将使您可以专注于改进核心产品。

More so, leveraging APIs can lead to several improvements across most crucial elements of your data science projects. According to a recent study, users who adopted APIs realized a significant increase in productivity (at 59%), direct revenues (at 43%), and innovation (at 51%).

更重要的是，利用API可以在您的数据科学项目的最关键要素上带来多项改进。 根据最近的一项研究 ，采用API的用户实现了生产率(59％)，直接收入(43％)和创新(51％)的显着提高。

Therefore, the ability to find and connect to APIs is an important feature you can take advantage of to increase the efficiency of your projects.

因此，查找并连接到API的能力是您可以利用的一项重要功能，可以提高项目的效率。

4.巩固API安全性 (4. Solidifies API security)

When using APIs in your data science applications, security should be an important consideration.

在数据科学应用程序中使用API​​时，安全性应该是一个重要的考虑因素。

Currently, with the increased utilization of APIs, attackers are shifting their focus from the traditional targets to a new area that hasn’t been extensively exploited yet: application programming interfaces.

当前，随着API利用率的提高，攻击者将重点从传统目标转移到了尚未被广泛利用的新领域：应用程序编程接口。

APIs are highly susceptible to attacks because they expose the underlying enactment of a computing technique.

API非常容易受到攻击，因为它们暴露了计算技术的基本实现。

Since they provide programmatic access to data and services, this also enlarges the attack surface area — unlike other user-facing applications.

因为它们提供对数据和服务的编程访问，所以这也扩大了攻击的范围-与其他面向用户的应用程序不同。

According to a recent report, the number of API vulnerabilities has been rising tremendously year-over-year, increasing by about 154% from 2015 to 2018.

根据最近的一份报告 ，API漏洞的数量同比激增，从2015年到2018年增加了约154％。

By managing APIs, you can enforce their security and realize the value you initially envisioned with your data science projects.

通过管理API，您可以加强其安全性并实现您最初在数据科学项目中设想的价值。

A management platform can help you to maximize your API’s security from various types of attacks, such as man-in-the-middle, denial of service (DoS), injection, or social engineering attacks.

管理平台可以帮助您从各种类型的攻击(例如中间人攻击，拒绝服务(DoS)，注入或社会工程攻击)中最大化API的安全性 。

Besides authenticating and authorizing access, an API management tool can assist in setting up various security measures for your APIs.

除了对访问进行身份验证和授权外，API管理工具还可以帮助您为API设置各种安全措施。

These include configuring the OAuth token authorizing system, instituting policies to block bad actors, revoking compromised APIs, implementing API key rotation, applying quotas and rate limiting, and setting permissions for accessing APIs.

这些措施包括配置OAuth令牌授权系统，制定策略以阻止不良行为者，撤销受侵害的API，实施API密钥轮换，应用配额和速率限制以及设置访问API的权限。

Also, as earlier mentioned, such a platform will increase the visibility into your APIs. This heightened visibility lets you analyze the API activities against normal usage patterns, discover unauthorized intrusions based on strange behaviors, and identify threats as early as possible before the extent of damage is amplified.

另外，如前所述，这样的平台将增加对API的可见性。 这种增强的可见性使您可以对照正常使用模式分析API活动，基于奇怪的行为发现未经授权的入侵，并在损害程度扩大之前尽早识别威胁。

5.启用API货币化 (5. Enables API monetization)

API monetization, just as the name suggests, refers to the process of earning revenues from your APIs. Monetization allows API providers to grow their pursuits and discover new opportunities with developers, partners, and customers.

顾名思义，API货币化是指从您的API中获得收入的过程。 货币化使API提供者可以不断发展追求，并与开发人员，合作伙伴和客户一起发现新的机会。

Depending on your objectives, there are three main monetization options you can pursue with your API program: providing it for free, the consumer pays, or consumer gets paid.

根据您的目标，您可以使用API​​程序进行三种主要的货币化选择：免费提供，消费者付费或消费者付费。

For example, if you have created an algorithm that performs sentiment analysis, you can expose it as an API and allow other developers to use it to find the opinion expressed in documents or articles.

例如，如果您创建了一种执行情感分析的算法，则可以将其公开为API，并允许其他开发人员使用它来查找文档或文章中表达的观点。

And to generate some money from the use of your sentiment analysis API, you can monetize it by charging other data scientists for using it.

为了通过使用情绪分析API来赚钱，您可以向其他数据科学家收取使用费，从而从中获利。

Most API management solutions offer an easy way to monetize and unlock value from your API products.

大多数API管理解决方案都提供了一种简便的方法来从您的API产品中获利和解锁价值。

These platforms will provide you with a centralized payment center for your APIs, the ability to implement subscription plans, billing capabilities, and more.

这些平台将为您提供API的集中式支付中心，实施订阅计划的能力，计费功能等等。

结论 (Conclusion)

APIs are widely used in data science projects. However, ad hoc API adoption often leads to performance discrepancies, security issues, and reduced productivity.

API已广泛用于数据科学项目中。 但是，临时采用API通常会导致性能差异，安全性问题和生产率下降。

An API management plan provides a single source of truth to bring unity, cohesion, and versatility to your APIs.

API管理计划提供了一个真实的来源，可为您的API带来统一性，凝聚力和多功能性。

With a platform to manage and control your APIs, you can optimize their value and make the most of your data science projects.

借助用于管理和控制API的平台，您可以优化其价值并充分利用您的数据科学项目。

So, are you using (or planning to use) a tool for the management of APIs?

那么，您是否正在使用(或计划使用)用于API管理的工具？

Or, do you have any comments or questions?

或者，您有任何意见或问题吗？

Please post them below.

请在下面发布它们。