Holen Sie sich die neuesten Updates von Hortonworks per E-Mail

Einmal monatlich erhalten Sie die neuesten Erkenntnisse, Trends und Analysen sowie Fachwissen zu Big Data.


Sign up for the Developers Newsletter

Einmal monatlich erhalten Sie die neuesten Erkenntnisse, Trends und Analysen sowie Fachwissen zu Big Data.


Erste Schritte


Sind Sie bereit?

Sandbox herunterladen

Wie können wir Ihnen helfen?

* Ich habe verstanden, dass ich mich jederzeit abmelden kann. Außerdem akzeptiere ich die weiteren Angaben in der Datenschutzrichtlinie von Hortonworks.
SchließenSchaltfläche „Schließen“
January 04, 2018
Vorige FolieNächste Folie

Applying Big Data Streaming Analytics in the Real World

IoT, the Internet of Things, has been a buzzword for the past five years. Literally everyone across all industries – business executives, line of business owners, operation staff, mechanical engineers, even retail marketers – has been eyeing the benefits and market opportunities IoT initiatives can bring. Yet there are so many points of view on IoT strategy that it is tough to have an agreed-upon, singular conversation on the topic. My definition of an IoT strategy simply means the actions based on the insights extrapolated from the traversing streams of data generated by connected endpoint devices, such as sensors, mobile phones, work stations, wearable devices and more.

However, with the speed of business today, enterprises are no longer satisfied with an IoT solution without real-time capability. The real game-changer is the ability to collect, curate, aggregate, analyze, and act on data, in real-time, such as that offered by Hortonworks DataFlow (HDF™). Hortonworks DataFlow is a data-in-motion platform to build dataflow management and streaming analytics applications, across the data centers, cloud, edge devices and sensors. With HDF, customers are able to exploit the advancement of big data streaming analytics to enable right-time, data-smart decisions.

Additionally, organizations can gain an edge to outpace the competition by grasping customer sentiments that align with the latest market trends, innovate faster by building products and services tailor-made to the changing market, and formulate strategies that capitalize on new venture opportunities. To further demonstrate the capability of HDF, the following solution use cases showcase the utilizations of the platform and their business benefits in the real-world scenarios:

Ad Server – what is an ad server? Here is a detailed explanation: In a nutshell, it is a web-based technology that employs a real-time bidding and auction process (often done in milliseconds) to help publishers efficiently manage all of the ad space on their sites, as demonstrated in the chart below:




With HDF, Hortonworks introduces a new approach to ad serving technology to keep pace with the evolving consumer and market needs. This 100% Open Source technology integrates disparate data sources with Machine Learning for better discovery and tracking of ad buying experience. Capabilities of the solution include:

  • Real-time enrichment of click-stream with internal data, such as demographics, behaviors, etc.
  • Apply predictive behavior modeling to enhance rule-based models
  • Centralized administration and monitoring, on premise or in the cloud

Based on the contextual analytics insights, business users can then ask new questions, measure campaign results across a variety of devices and formats, or adjust their customer engagement strategy to optimize revenue channels.

To learn more about Hortonworks solution for Ad Server, please visit:
Ad Server Solution Sheet
Case Study

Clickstream Analysis – clickstream data is an information trail a user leaves behind while visiting a website. It is typically captured in semi-structured website log files. These website log files contain data elements such as a date and time stamp, the visitor’s IP address, the URLs of the pages visited, and a user ID that uniquely identifies the user. As we are already in a Digital Age where purchasing and gathering of information are done on web – roughly 8 in 10 Americans are online shoppers based on a recent study – the analysis of these web logs becomes crucial to the success of a website as well as a business entity. Take Amazon for example, an estimated 35% of Amazon’s product sales comes from product recommendation, and how it is done? Through the power of clickstream analytics capturing and analyzing the digital footprints of millions to billions of navigation clicks per hour or day.

The Hortonworks Clickstream solution utilizes a combination of the Hortonworks Data Platform (HDP™), Hortonworks DataFlow, and our Professional Services to setup the implementation. Capabilities of the solution include:

  • Ingest Pipeline for consuming digital events, customer information and third-party sources
  • Transformations for user sessionization, segmentation, data cleansing, mapping and aggregation
  • Storage & Analytics of data to be accessed by multiple users and other systems
  • Visualization for presenting easy consumable insights

The insights gathered from this solution lead to tremendous business value, such as:

  • Enhanced visitor experience with easier navigation through products and services on the website
  • Maximized upselling opportunity by proposing the right product recommendations to increase basket size
  • Improved purchase conversion rate by influencing visitor behaviors with personalized marketing and promotions

To learn more about Hortonworks solution for Clickstream Analysis, please visit:
Clickstream Analysis Solution Sheet
Visualize Website Clickstream Data
Hadoop Tutorial: Analyzing Clickstream Data

Predictive Maintenance – predictive maintenance analytics is perhaps the most prominent use case of IoT, widely adopted in industries such as Manufacturing, Oil and Gas, Transportation, and Aerospace and Defense over the past years. One apparent explanation to its surge is that the solution proposes significant cost and time savings by analyzing equipment sensor data, diagnostics readings, and maintenance records to predict which components, at the approximate time period, will fail. Rather than a traditional break-and-fix approach that results in expensive repairs and lengthy equipment downtime, businesses can now take preventive measures based on the predicted failures to reduce maintenance costs while improving equipment uptime before an outage occurs.

The predictive maintenance solution relies on the robust data collection capability of HDF and the scalable data storage and discovery of HDP to move, process, and analyze sensor and diagnostics data for the training of predictive machine-learning models. Capabilities of the solution include:

  • Cost efficient, horizontally scalable storage unites all your sensor, diagnostic, and maintenance data in a central data lake
  • Real-time, easy to use, scalable and secure ingestion of sensor data from anywhere
  • Flexible solution tailored to your specific requirements and integrating with your processes and toolsInnovative open source platform with complete control and access to all your data
  • Flexibility to deploy to the cloud, on premise or hybrid cloud

This solution benefits enterprises in numerous ways; time and cost savings are just the tip of an iceberg. Enterprises across industries apply the solution to enhance safety, optimize maintenance staff, improve parts inventory to reduce under/over stock, and avoid environmental and production waste due to defects produced by failing machinery. Furthermore, as many industrial or heavy-machinery manufacturers transform their business models to incorporate more service offerings as additional revenue streams, the solution enables them to employ a proactive approach with accurate predictions of component failure to capitalize on maintenance and servicing opportunities.

To learn more about Hortonworks solution for Predictive Maintenance, please visit:
Predictive Maintenance Solution Sheet
Webinar: Predictive Maintenance for Automotive Industry

Parting thoughts
You can read more about our Hortonworks DataFlow platform here. Also, watch our recorded webinar, HDF 3.0 with Live Demo, with Haimo Liu, Sr. Product Manager at Hortonworks demonstrating how easy it is to build your own data flow and streaming apps.

Antwort verfassen

Ihre E-Mail-Adresse wird nicht veröffentlicht. Pflichtfelder sind mit * gekennzeichnet