The Failure and Success of Unstructured Data Reuse a Pragmatic Analysis

Structured data is easy to parse and process using simple software tools with effective results. However, most valuable information is typically found in unstructured data. The problem is that unstructured data is significantly harder to process, especially at large scale. Adding some structure can greatly improve data reuse. To demonstrate our approach, we focus on one particular example, namely the design patterns, a common encapsulation mechanism, widely used in several domains to capture and reuse valuable domain experiences. Patterns are growing in numbers as an effective way of communicating knowledge among expert designers. Currently, there are numerous domains where patterns are applied, including HCI Human Computer Interaction. HCI Patterns are typically published in books or on the Internet. The sheer number of HCI patterns, the rich and heterogeneous contents coupled with the lack of a delivery system can confuse and overwhelm a novice pattern user, even when they are meant to help novice users in the first place. We present an approach to add structure to HCI patterns, hence providing the ability to communicate and reuse them in a more scalable and effective way.