文章转载自微信公众号北斗智造者
编者荐语：
「“互联网+”时代，数据呈现多元化趋势，进行地理空间智能的研究受到广泛关注，其商业价值和社会价值有着无限可能性，并蕴含在大数据分析中。那么地理空间数据如何助力智能化进阶呢？」
大数据分析影响着我们日常生活的各个领域，也同时改变着地理空间数据的使用方式，而且这种影响是双向的。正如分析海量数据的能力使地理空间数据比以往任何时候都更强大、更有价值，物联网带来的地理空间智能也赋予了大数据分析巨大的动力。
今年，从物联网帮助应对新冠肺炎疫情，到人们预测通过实时物流监控和一系列相关技术改变电子商务现状，我们已经看到了很多这样的例子。
这些新的发展非常振奋人心。本文将介绍地理空间大数据分析的发展现状，地理空间智能与大数据在哪些方面有建设性的重叠，以及两者未来的发展前景。
01、大数据和GIS
首先要注意的是，地理空间数据和大数据分析的加速融合并不是一个全新的现象。追溯到2011年，麦肯锡已将其列为下一个创新前沿。并且在那个时候，我们就已经看到了一些具有创新性和交叉性的解决方案。
然而在该报告发表后的近十年里，硬件成本阻碍了地理空间数据收集设备的大规模部署；其中，成本占比最大的是存储成本。例如，在2010年，每千兆字节的计算机存储成本是10美分，而到2017年，每千兆节的存储成本变为2美分，为之前的五分之一。如今，随着数据储存成本的持续下降，即使是小公司也终于可以在地理空间数据基础上部署大数据分析。
硬件成本并不是地理空间数据分析所面临的唯一挑战。在过去的一年里，特别是在为抗击新冠病毒而部署的地理追踪，人们明显对收集他们相关活动的数据以及如何使用这些数据表现出了担忧。虽然向更私人的工具和应用程序的转型并不是一件容易的事，但它在某种程度上可能会对地理空间智能行业产生影响，毕竟，这种分析得出的预测仅与输入的数据有关。
02、应用
尽管存在这些挑战，但在未来几年，无论是在学术理论还是方法论上，地理空间数据和大数据分析的结合将会更加紧密。
最近的市场报告显示，地理空间智能领域将会出现大幅增长。在2020年，全球地理空间数据分析的市场收入从2018年的699亿美元增加到883亿美元。这不仅代表了行业内的一项重大投资，也证实了它正在成为技术经济中更安全、更稳定的一部分。
毋庸置疑，地理空间数据分析将在未来几年使用到新的应用中。如果要预测未来的发展趋势，我们有必要先了解一下它已经在哪些领域发挥作用。以下是三个主要应用领域：
1. 人道主义援助
地理空间大数据分析最重要的应用之一是在人道主义救援领域。现在世界各地都在用GIS物联网设备，在救援人员不易进入和难以工作的区域进行数据收集。
据DigitalGlobe分管总监Abhineet Jain介绍，该组织每天收集约80千兆字节的数据。截至2018年1月，该组织收集到的数据总量接近100PB。如果没有大数据技术，这个数据量是不可能完成的。
2. 市场营销
地理空间大数据分析在市场营销中的应用更为广泛，现在许多品牌都根据活动和位置追踪器的数据，来分析向客户提供产品的范围。这依赖于机器学习系统，并且大数据革命已经助力机器学习系统成为主流。
举个例子，Under Armour公司利用健身追踪器的数据对健身人士根据运动水平进行划分，甚至为经常参加各种运动的客户量身定制产品。
3. 商业智能
几年前，人们很难想象金融部门和地理空间数据是如何协同工作的。因为对于银行或其他金融服务公司来说，了解客户的旅行时间和地点似乎没有什么价值。
事实证明，这些数据对金融业和其他行业一样有用。事实上，金融领域的地理空间大数据在当前的创业热潮中扮演着重要角色，旨在将地理空间分析技术纳入商业决策的核心。
地理空间大数据分析的应用仍在探索中，但已经看起来非常有前景。地理空间数据的应用已经十分广泛，例如，确定哪些分支机构需要合并，以及如何利用卫星图像更好地预测一段时间内的洪灾风险以确定保险费率。
03、未来预测
尽管涉及大数据和地理空间数据融合的领域仍在发展初期，但不难预测其发展方向：这两个学科有很多东西可以相互借鉴，在未来十年可能会变得越来越难以区分；再加上新技术，尤其是5G在物联网中的作用日益显著，几乎可以肯定的是，我们正处于这两个领域革新的起点上。
另附原文，供大家参考：
How geospatial intelligence powers Big Data analysis
by Bernard Brode
Big Data analysis has impacted almost every sector of our economy, so it’s no surprise that it is also transforming the way that we work with geospatial data. This impact is a two-way, though. Just as the ability to analyze more data than ever before is making geospatial data more powerful and valuable than ever before, geospatial intelligence drawn from the IoT is super-charging Big Data analytics.
This year, we’ve seen plenty of examples of this, from the way that the IoT has helped in the Covid-19 fight to speculation about the way that it has changed ecommerce via real-time logistics monitoring and a host of associated technologies.
With these exciting new developments in mind, in this article we’ll take a look at the current state of the art when it comes to geospatial data in Big Data analysis, the factors that drive the current boom, where the two fields have constructively overlapped, and what the future holds for both.
01
Big Data and GIS
The first thing to note about the quickening merge of geospatial data and Big Data analysis is that this is not an entirely new phenomena. McKinsey highlighted it as the next frontier for innovation way back in 2011. Even at that near “pre-historical” time we were seeing some innovative, cross-over solutions that drew on both fields.
For almost a decade after that report, however, the cost of hardware held back mass deployment of devices that could collect geospatial data. One of the largest components of this cost was that of storage. For example, the cost per gigabyte for computer storage in 2010 was 10 cents. In 2017, that dropped by a factor of five to two cents per gigabyte. Today, as these costs continue to fall, it has finally become feasible for even small firms to deploy Big Data analysis on geospatial data.
Not that this is the only challenge facing this kind of analysis. In the past year, and particularly in the context of the geographical tracking that has been deployed to fight the Covid-19 virus, citizens have raised legitimate concerns about the amount of data collected on their movements and how it is used. While the shift to more private tools and applications isn’t a headlong rush yet, at some point it may have an affect on the geospatial intelligence industry. After all, the predictions such analysis yields is only as good as the data going in.
02
The applications
Despite these challenges, it’s likely that geospatial data and Big Data analytics are likely to draw closer together – both as academic disciplines and methodologies – in the coming years.
In fact, recent market reports point to dramatic growth in the sector. The global geospatial data analytics market is expected to increase in revenue from $69.9 billion in 2018 to $88.3 billion in 2020. Not only does this represent a significant investment in the industry, but also illustrates the fact that it is becoming a more secure, stable part of the technology economy.
While geospatial data analytics will doubtless find new applications in the coming few years, the largest growth will come – at least initially – in the sectors and applications where it is already in use. For clues as to the development of the sector, it’s therefore worth looking at where it is strong already. This is in three main fields:
1. Humanitarian aid
One of the foremost applications of geospatial Big Data analytics has been in the humanitarian sector. GIS IoT devices are now being used across the world to collect data in environments which were previously difficult for aid workers to access and consequently difficult to work in.
For an example of the way in which geospatial Big Data analytics can work well in this sector, take a look at the work of DigitalGlobe, a non-profit organization that sources satellite data and integrates it with other sources like social media sentiment and aerial imagery, leverages a GIS machine learning algorithm to track activity in specific locations and identify anomalies.
According to DigitalGlobe regional director Abhineet Jain, the organization collects approximately 80 gigabytes of data daily; as of January 2018, the organization had collected close to 100 petabytes of data total, an amount that would be impossible to work with if it wasn’t for Big Data techniques.
2. Marketing
A more widespread use of geospatial Big Data analytics has been in marketing. Many brands now use data drawn from activity and location trackers to inform the range of products they offer to customers. This type of analysis is dependent on the kind of machine learning systems that the Big Data revolution has helped bring into the mainstream.
As an example, it’s worth reading about the way that Under Armour uses data from fitness trackers to segment their audiences based on their level of physical activity and even tailors product recommendations to customers who regularly engage in various types of sport.
3. Business Intelligence
A few years ago, it was difficult to imagine how the financial sector and geospatial data would work together – there appeared to be little value to a bank or other financial services company in knowing where their customers traveled and when.
As it turns out, this data is just as useful to the financial sector as it is in other industries. In fact, geospatial Big Data in the financial sector now plays a role in the ongoing startup boom that aims to bring geospatial analysis techniques to the heart of business decisions.
The applications are still being explored but already seem promising. Geospatial data has already been useful, for instance, in determining which branches to consolidate, as well as how satellite imagery over time can better predict a property’s risk of flooding when it comes time to determine insurance rates.
04
The Future
Though the confluence of Big Data and geospatial data is still relatively young, it’s not hard to predict the direction of travel: the two disciplines have much to learn from each other and are likely to become increasingly indistinguishable in the next decade. Add to this the emergence of new technologies – and particularly the role of 5G in the IoT – and it seems almost certain that we are on the threshold of a revolution for both fields.
转自：“量化研究方法”微信公众号
如有侵权，请联系本站删除！
