2019年是知识图谱相关技术飞速发展的一年，世界顶级NLP大会ACL一次性收录了超30 篇高质量知识图谱类论文，其中对于关系向量和图神经网络的论述将对知识图谱中关系补 全、推理决策和认知计算带来长足发展。在国内建设应用方面，工信部发文明确指出， 2020年将围绕工业大数据融合应用、民生大数据创新应用、大数据关键技术先导应用、大 数据管理能力提升4大类7个细分方向着重发展，而知识图谱作为集大数据和人工智能与一 身的综合技术，也将成为重点关注领域。

The key objective of big data analytics system is to help organizations get a better understanding of their customers and help it narrow down their targeted audience; thus, improving companies marketing campaign. Numerous benefits of big data analytics such as allowing customers to capture and process different services, modify existing orders, process customer moves and price quotes to the customers. The main purpose of big data analytics solution is to verify and process the telecom order based on its product catalog and service catalog. Rise in demand for network installations across rural areas and rapid increase in connectivity devices and subscribers are the major factors that drive the growth of the Big data and business analytics market. Furthermore, integration of advance technology such AI, machine learning and big data and lack of standardization and compatibility issues with the existing systems are anticipated to provide lucrative opportunity for the market.

The data center industry is witnessing rapid growth and is focusing more on efficiency and maximum uptime. For instance, according to the CloudScene, there are 2,539 and 428 data centers in the United States and Germany, respectively (as of Q4 2019). With increasing capacity and higher density, there is an increasing need for energy-efficient cooling of the IT equipment. Cooling remains as one of the major challenges faced by data center managers, as it accounts for around 20% of the overall expenditure. Air and liquid cooling are two of the most commonly used methods for heat management in the data centers. While air cooling is less pricey and messy, as compared to liquid cooling, liquid cooling is considered more beneficial over air cooling, as water is more efficient than air as a medium of heat removal and can reduce cooling power needs by 70%.

A major driver for the adoption of digital asset management (DAM) solutions is the high return on investments. As a result, the spending by organizations for asset management is expected to increase, during the forecast period.Customer engagement is emerging as the most effective method of customer retention, and mobile phones are the foundation of this trend. The number of connected devices is increasing, globally. With solutions, like interactive store windows, digital signage, wearable devices, marketers are finding it difficult to create and deliver assets that offer consistent and compelling brand messages across the screens. These marketers must develop content for diverse audiences, which is either automatically responsive (resizable) or completely reconfigured for different screens. As a result, marketers are investing in DAM solutions to help organize and transform their assets, automatically, into the right size and formats to reach customers, ever ywhere and on any screen.

数据爆发增长及企业上云需求带动数据中心基础建设不断扩张，指数级增长的数据流量和线性增长的IDC机房规模推动行业进化（超大型云化数据中心和边缘数据中心）。想要提高数据中心存储和计算能力，可以从三个方面考虑（性能、利用率和规模数量）。由于考虑摩尔定律减速，性能提速逐渐达到极限；利用率目前主要利用虚拟化技术，企业想要提升数据中心运算能力只能从数量入手，因此扩大单体规模来提升整体效能成为有效解决方式之一。此外，随着5G网络建设启动，伴随uRLLC低时延应用场景的部署，包括物联网、车联网、工业互联网等系列推进，对小型边缘IDC需求有望增长。

VR 四大技术特征解读，5G 时代交互与硬件限制有望得到解放。我们总结出限制VR 设备普及率的四大技术因素：1、设备重量与性能的平衡，目前VR 一体机虽缓解设备重量问题，但由于性能较差用户体验得不到提升；2、显示方面，光学元件能否解决晕眩感、纱窗效应、提高沉浸感至关重要，AMOLED 能有效降低时延从而缓解晕眩感，但价格高昂；3、交互方面，Inside-Out 的6DOF 技术+虚拟移动技术可解决VR 游戏领域移动交互与视觉内容适配的问题；4、芯片方面，目前国内外主流VR 头显设备主要采用高通骁龙835 芯片，芯片性能相比于桌面级CPU 与GPU 仍有较大差距，导致VR一体机用户体验较差。我们判断，随着5G 时代到来，大带宽、低时延将促使VR 交互与硬件限制得到解放，控制成本的前提下提升用户体验，高品质VR 产品有望得到普及。

从我国第三方移动支付市场的发展历程来看，根据不同时期的主要增长点不同大致可以分为三个阶段。第一个阶段是 2013-2017年的线上场景驱动阶段，电商、互金、转账的先后爆发持续推动了移动支付的快速增长。第二个阶段是2017- 2019年的线下场景驱动阶段，2017年开始线下扫码支付规模全面爆发增长，线下场景的支付增速远高于线上场景支付的 增速，引领移动支付经历了由线上驱动阶段到线下驱动阶段的转变。第三个阶段是从2019年开始的产业支付驱动阶段，以 C端驱动的线上线下支付因C端流量见顶都进入了平稳增长期，而产业支付伴随产业互联网的快速崛起正逐渐成为我国移动 支付新的增长点。