Enterprise Swarm

We highlight three use cases in which big data may create value by increasing organizational productivity:

As employees perform their day to day activities, they generate a myriad of textual data, including documents, email, presentations, patents, blogs, chats, tweets, etc. The corpus of these textual data contains vast untapped organizational knowledge. The Connected Enterprise emerges when we combine the wealth of tacit knowledge conveyed in organizational document corpora with enterprise social connections. Advanced analytics over the combined information becomes the first aid for productivity, by inventing algorithms for content-based contextual search, expert suggestion, understanding an organization's expertise, finding gaps in organizational knowledge, creating work teams, and more.

Modern physical systems are outfitted with sensors, which generate a stream of data about processes occurring in the system, much like system logs contain ongoing data about software systems. Expert engineers often use such data to troubleshoot problems in R&D or for customer service, but capturing such expert knowledge is complex and time consuming. We are creating algorithms and methodologies that extract insight from data with little or no expert intervention. These analytics can transform sensor and event data to monitors, baselines, and patterns, which can, in turn, inform about system health and fleet performance so that systems may be repaired faster and maintained proactively.

At the end of the day, applications are constructed to deliver functionality to users. Maximizing business outcomes and improving the user's experience go hand in hand. We need to understand the user in order to deliver effective and efficient experience. By analyzing the behavior of real users, we propose to make the most relevant content available, whether it is news or advertisements. We may rearrange content and links to render the next step in a business process easily accessible. User analytics can even drive the process of application design and enhancement. Borrowing methods from sequence analysis and collaborative filtering and extending those approaches to the complexities of user behavior analysis scenarios, we are inventing methods to understand what users are doing and how to enable them to do it quickly and easily.

• HP Collective
• Regex GPU Library
• Automatic Policy Extraction