OpenAmplify is proud to release Version 2.0 to the world’s first comprehensive semantic platform. Full of new features, but still compatible with V1.1, OpenAmplify 2.0 reflects our commitment to delivering real-world, groundbreaking advances to the Semantic Web community.

As part of our latest release, OpenAmplify version 2.0, we offered a live webinar. OpenAmplify CIO Mike Petit spoke about this new release and answered attendee questions.

• PinPoint: a new topic scoring algorithm that helps surface the topics with which the author is most “engaged.” Which topic is most important? Which carries the most sentiment, or is the subject of the most guidance? About which topic is the author most likely to actually do something? All of these questions, once answered, tell us which topic is truly the most important in the text. PinPoint delivers this sophisticated combination of signal analysis in a simple manner. Adding “scoring=pinpoint” to your 2.0 API call is all you need to do. Topics are returned as they always have been, but with PinPoint weighting applied to their scores. PinPoint is a true breakthrough in keyword identification;

• Contrast:a brand new signal, part of the Style analysis. Contrast measures the degree to which disagreement is detected in the text. This should be of great value to CRM systems, moderation applications, law enforcement and other applications where understanding the “heat” is important;

• Named Entity Recognition: as a precursor to an important feature in a later release (called co-reference; ask us in the forums and we’ll explain it), we now identify in the Topic analysis any recognized named entities (companies, persons and places). We are also much better at recognizing multi-token NE’s such as “University of North Carolina” or “Department of Education”, which should improve your applications considerably. Subtype information is also provided when possible, such as male or female for persons;

• Text-level Polarity, Guidance, Decisiveness and Temporality: we now include these signals at the level of the full text, in the Styles analysis. This is of particular value in short texts like tweets, where there may not be an identifiable topic (“that’s awesome” has no topic, but definitely has polarity). These new signals should help those following threads, or measuring sentiment in a known brand context, to catch more of the vibe in topic-starved content.

There’s more; please refer to the documentation for details.

OpenAmplify Version 2.0 is the result of many months of hard work, driven by needs expressed by you, the Community. We hope you’ll find it as exciting as we do. Dive in!