Friday May 06, 2016

Smart user experiences: Machine learning and the future of enterprise applications

By Bill Kraus, Oracle Applications User Experience

Self-driving cars, drones that will deliver packages to our doorsteps, household robots that express emotion, and even the fear that runaway artificial intelligence will enable machines to enslave us and become our overlords – these examples of machine intelligence demonstrate how science fiction is rapidly becoming science fact. But while this all makes great fodder for Internet memes, why should you care? How could machine intelligence affect your bottom line in the enterprise world?

To answer this question, the Oracle Applications User Experience (OAUX) team has been exploring ways to leverage  machine intelligence as part of our Glance, Scan, Commit design philosophy and strategic user experience themes of Simplicity, Mobility, Extensibility, which are already changing the way you do your work.  

But first, we need to make a distinction between the underlying technology, and the way we interact with it. The combination of machine learning – where computers can learn without being explicitly programmed,  and Big Data – massively large datasets that can be analyzed to reveal subtle but significant patterns and trends, has dramatically enhanced our ability to build systems that recognize, classify, and analyze the world around us. This approach is currently being used for everything from facial recognition to online shopping recommendation systems. 

As interesting as this advancement is, its potential can only be truly leveraged through an intuitive, contextual user experience that anticipates a user’s needs and goals –  something we are calling smart user experiences. A smart user experience is not tied to any particular underlying technology, but rather uses whatever technology is appropriate, regardless of whether it is the tech du jour, to create a user experience that just works.

Focusing on the technology alone misses a critical point that is often overlooked – no matter how ”cool” or cutting edge a particular technology is, its ultimate utility comes from how easily we can integrate it into our daily routines. In fact, many of the technologies that have transformed our lives have done so because they have made access to technology easier. For example, while advances in mobile technology have provided the technical foundation, it is the simple, intuitive, gesture-based user experience that has allowed smart phones to transform how we work.

There are three ways in which smart user experiences can have a positive and significant impact on our professional and personal lives: through automation, by offering advice, and by enhancing exploration and discovery. 


The first advantage of smart user experiences is to automate the mundane, that is, automatically perform the perfunctory, day-to-day tasks and offload the myriad of prosaic decisions we face everyday so we don't have to. The line between work and play has nothing to do with the amount of energy we exert. Most of us likely exert more energy in our leisure activities than we do sitting at a desk. But the perception of what is work and what is play is directly tied to the level of drudgery we experience. By being able to delegate humdrum tasks to the system, our businesses can not only be more efficient, but we can focus on those creative aspects of our work. 

Examples include everything from auto-filling forms to speeding navigation to sending automatic notifications to prioritizing and performing tasks, all day-to-day tasks where smart user experiences can leverage context and require minimal oversight.


Recommendations from online retailers are an obvious example of advice, but this is just the tip of the proverbial iceberg. 

Smart user experiences not only recommend information and opportunities that might be interesting to us, but can guide us through complex transactions, advising us on the best course of action to achieve our business objectives.


Smart user experiences also can help us discover new opportunities and make connections that we would have otherwise missed. Machine intelligence can analyze thousands of variables across terabytes of data, potentially uncovering subtle but significant relationships. Yet this complexity quickly becomes overwhelming, risking paralysis by analysis. Smart user experiences are designed to offer information to users in a way that is both digestible and actionable while simultaneously inviting exploration –  something that we humans enjoy and are very good at. 

All three advantages have the potential to fundamentally change the way we do business by not only making us more efficient, but also more effective at finding and exploiting new business opportunities. 

Fostering the Human Machine Dialog

Far from being marginalized, the advent of intelligent systems has made the user experience all the more important. No longer just “dumb” tools, the technology with which we interact will soon be able to help save time by speeding up repetitive tasks, enhance our competitive edge by supporting our decision- making, and in some cases, autonomously carry out tasks. Our flint arrowheads and stone hammers have grown into virtual personal assistants and advisers.   

User experience in this context means first understanding which approach will best support what a user needs to do, providing the user with equal parts control and automation while acknowledging that different users expect different levels of oversight. Smart user experiences recognize this, building environments that foster a nuanced dialog between the two complementary but highly interdependent forms of intelligence: human and machine.  

In the movie 'The Usual Suspects,” Kevin Spacey's character has the memorable line: ”The greatest trick the devil ever pulled was convincing the world he did not exist." Similarly, the best user experiences are those that are so natural, so intuitive, so seamless, that they are all but transparent. 

Smart user experiences provide the means to achieve this, and they are a natural extension of the Glance, Scan, Commit design philosophy that has transformed the Oracle Applications Cloud user experience. It is about understanding how we interact with machine intelligence in a way that is comfortable to us, leveraging context and computational resources to enhance our knowledge and enterprise situational awareness.

In a nutshell, smart user experiences offer a more human way of working.

Stay tuned as we explore the topic of smart user experiences and how Oracle is incorporating them in upcoming posts on the VoX blog. 

Friday Aug 16, 2013

Emerging Design Principles for End User Consumption of Big Data

By Misha Vaughan, Oracle Applications User Experience

Editor’s Note: This is part 3 in a three-part series on the user experiences of working with big data. In the last post on this topic, John Fuller, Consulting User Experience Designer for Endeca, wrote about some of his team’s key requirements for designing usability into the user interfaces for Endeca Information Discovery. In this post, the emerging thinking on design principles for delivering all this power to regular end users is the topic. Thank you to peers John Fuller, Julia Blyumen, Edward Roske (@eroske), and Aylin Uysal for the inspiration of these themes.

Information visualization is a whole field unto itself, and education is now widely available on this topic, notably Edward Tufte’s work on Information Visualization.

When information visualization was discussed at a recent summit on user experience for big data summit, a specific new insight for me was that I saw a set of information visualization guidelines emerging for end users. I don’t mean data analysts or business analysts who are doing deep, big data analysis.  I mean the end user, for whom the analyst is preparing data.  

How do you present big data to an executive or a decision-maker in a way that is digestible? How do you take them from the big picture insight, down into the supporting details? Do you show them the trellis charts and say “see here?” Or do you take a more narrative approach?

In no particular order, these were my lessons learned about end user design principles for big data visualizations:

1.    Make the invisible visible.

The entry into a big data analysis can be through seemingly simple information visualizations. Take a strategy from the newspaper industry’s use of infographics, such as the Huffington Post or USA Today.  Through visualization, you can help the user better connect and interact with the data.  Information visualization and infographics are a core part of making the results of big data accessible.

2. Show the forest, then the trees. This is also known as progressive disclosure.

With more and more data available in larger amounts, end users now need, more than ever, attention to how to cleverly and conveniently discover what they need to know. Then they need to be given the ability to explore that data.

3.  It’s all about me, or staying in context of my task.

Making big data relevant to end users means considering how to display large quantities of data in the context of different enterprise use cases, such as human resources processes, financial processes, or sales processes.  This can be any kind of data, whether it's pulling in transactional data, analytics, or social feeds.

4. Tell me a story.

Big data is, well, a lot of data. Providing narrative sources can add context and clarity to complex data. Doing this in a systematized way has even more interesting implications for enterprise use cases.

5. Make it mobile.

This one is kind of a no-brainer.  This is about giving end users the ability to make this kind of data available on tablet-sized devices.

6.  I can trust this, by you showing me how you got here.

Because of the complexity of the data, and the possible multiplicity of data sources, the ability to create confidence in the quality and the timeliness of the data are key to the experience.  It also means showing the path or way an analyst arrived at a particular conclusion.

7. Make it fun to play with.

One of the delightful characteristics of big data is that there really is a lot of data you can play with.  There is a sweet spot for the developer or designer who invents clever components that allow for the creative display and manipulation of complex levels of data.

8. One UI to rule them all.

End users don’t really care how many data sources you are bringing together. They just want the result. The best experiences will unify many data sources, transparently --  whether it’s Endeca, a data warehouse, or social feeds -- into one representation.

Again, I can’t claim credit for the concepts. I’m just summarizing what I learned on that day. If you want to see what this all means for Oracle Applications User Experiences, stay tuned and see what’s coming at OpenWorld 2013 this year.

Tuesday Jul 30, 2013

Key User Experience Design Principles for working with Big Data

By John Fuller, Consulting User Experience Designer, Oracle

Editor’s Note: This is part 2  in a three-part blog series on the user experiences of working with big data. In my last blog on this topic, I summarized the conversation from a one-day summit with a few key partners on the user experience landscape with big data.  In this blog, John Fuller, full time interaction designer for Endeca, shares some of his team’s key requirements for designing usability into the user interfaces for Endeca Information Discovery.

John Fuller
John Fuller, Consulting User Experience Designer for Endeca

About two years ago, we took look at the product we had and felt that there was a lot of opportunity that was, in many ways, fairly unique in the marketplace. It was at that point that we developed a set of core design principles to guide us in our work going forward.

We crystalized the things we thought were working well and sought to maintain that focus going forward. I wouldn't say that they were designed specifically with "big data" as the main focus; the principles are much more broadly applicable. We're focusing on helping people bring together a variety of data types in a fast and flexible way with lower cost, so from that perspective, we're targeting a really interesting part of the big data story.

Endeca healthcare demo
Screen shot from an Oracle Endeca Healthcare Demo showing how big data can guide the detection of healthcare problems.

We came up with 6 core design principles and details about what each one meant. One of the really interesting outcomes of this has been that the principles have really held up over time.

Here are the six core principles:
  1. Enhance Insight - The value of discovery tools lies in the insights they help discovery workers realize, by enhancing the natural ability of people to understand the answers that are in the data.
  2. Encourage Exploration - Discovery applications encourage exploration.  Users will want to ask new questions, pursue new avenues of exploration, and consider new connections and relationships across the diverse types of information presented by discovery applications.
  3. Coherence and Clarity - All elements of the experience should work together in a coordinated fashion. The way the system works is clear at all levels, making the results and implications of actions easy to understand.
  4. Readily Composable and Manageable - Creating, configuring, and managing discovery applications is straightforward and efficient. The product provides useful defaults, intelligent starting points, and encourages application builders to make good choices when composing discovery applications.
  5. Engaging and Compelling - Working with the tool is enjoyable, engaging, and satisfying, for new and veteran users. Endeca Information Discovery embodies and personifies the values and principles identified herein.
  6. Offer a Modern Application Experience - Discovery solutions “walk and “talk” like modern applications.
With every new project that comes up, the principles still seem new and fresh -- with new takeaways to guide the process. We're planning on adding more detail about the principles -- and several other topics -- on our blog, so if you'd like to hear more, check it out.

Tuesday Jul 23, 2013

The User Experience of Big Data in Oracle Enterprise Applications: Part 1 of 3

By Misha Vaughan, Oracle Applications User Experience

Editor’s Note: This is the first part of a three-part series on lessons we have learned about the user experience of big data, and trends in Oracle’s approach to the challenges of working with big data.

Misha Vaughan
Misha Vaughan, Director, Communications & Outreach, Applications User Expeirence
by Martin Taylor

I recently hosted a partner summit on the user experiences of big data at Oracle headquarters in Redwood Shores, Calif. The title of the summit was: “So You Have Big Data, Now What?”

The goals of the exchange were three-fold:
  • Assess where some key Oracle user experience partners -- Floyd Teter of EiS Technologies (@fteter), Edward Roske of interRel (@eroske), Mike Rulf of Core Services, and Ron Batra of AT&T (@ronbatra)-- were at in their conversations around the user experience needs of big data with their customers.
  • Discuss and sharpen our common understanding of the UX value propositions of some Oracle applications for big data. My particular interest was with OBIEE’s new information visualizations and Endeca Information Discovery’s UX.
  • Get feedback on a selection of forward-looking applications user experience innovation projects that intersect with big data. 
Below are my lessons learned from the conversation. Part 2, the next post in this series, is an email conversation with John Fuller, User Experience Designer for Endeca, on the key elements of designing user experiences for data analysts working with big data tools. Part 3 is a summary of what I see as the key UX design principles emerging in Oracle for a new class of design problems - making big data accessible to non-data analysts.

My Lessons Learned

Lesson 1: What customers are asking about “big data” and how they defining “big data”.

The general consensus was that some customers have already defined their strategy and are moving forward.  However, many customers are still trying to wrap their heads around what big data means for their institutions.  Our key partners see their customers’ understandings ranging across the following:
•    Big data is a massively large volume of structured data.
•    Big data is making sense of unstructured data, like Twitter feeds and Google search results (e.g., monitoring potential flu outbreaks).
•    Big data is about consolidating multiple sources of data, structured and unstructured, into one representation.
•    Big data is about solving wicked problems, for example, how to optimize something as complex as thinning a forest against needed output, aesthetics, and uncertain markets.
•    It is about discovering unlikely relationships in a large volume of data.

Lesson 2:  The big-data analyst is a highly specialized user role, and really needs the right user experience to be able to deliver the results companies are looking for.

Companies like Oracle are building the tools necessary for data analysts, such as Endeca's Information Discovery Tool.  Color me "wow" after seeing a demo by John Fuller.  Important tools in the toolkit are also OBIEE's "big data" visual analysis tools (thank you, Edward Roske).

This was a jam-packed conversation, and had so much in it that I decided to follow up with John and see if he would unpack the user experience requirements in more detail in a follow-up post. So stay tuned for that.

Lesson 3: It seems that there are really two user profiles we need to be concerned with in big data: the data analyst and the downstream producer, or possibly business analyst.

A recent study in the Wall Street Journal states that one of the biggest challenges of big data is finding professionals actually trained in the domain to help companies take advantage of this space. We know that the big business schools with IT programs will take the bait, but even that will not produce them fast enough. The rate of information is growing faster than our ability to sift it.

To take advantage of the sizeable investment required for a Big Data Project, a data analyst needs to enable a larger set of producers to leverage their data and share it with a larger audience. This may be a business analyst, or some other job title - but essentially this is a person who works with a lead data analyst to create the stories, visualizations, and associated analyses needed to communicate findings to a larger audience, which allows that lead analyst to get onto the next problem.

In my next post, I’ll write about Endeca, and the key elements of designing user experiences for data analysts working with big data tools.


Check here for news and upcoming events from Oracle's Applications User Experience team on the Oracle Applications Cloud and more.

People in Spheres
Misha Vaughan, Editor & Senior Director, Applications User Experience
@mishavaughan on Twitter

Learn more about us at
Usable Apps


« July 2016