About us
The Garden State Java User Group, Inc. (formerly the ACGNJ Java Users Group), in continuous operation since February 2001, is a 501(c)(3) non-profit organization that provides a forum to promote the use of Java platforms, APIs, and language. We strive to facilitate interactive discussions and exchange of ideas relating to the application of Java.
Topics will primarily focus on JVM technologies such as Java EE, Jakarta EE, MicroProfile, Spring, Groovy and Kotlin. Topics on non-JVM technologies such as Meteor, AngularJS, and Rust may occasionally be presented.
We are proud members of the MicroProfile Working Group and the Jakarta EE Ambassadors and have adopted the Jakarta NoSQL specification.
All meetings are generally held on the second Tuesday of the month, but may change due to speaker availability. In-person meetings are usually held at Drew University in Madison, New Jersey. On-line meetings are held via our YouTube channel.
Upcoming events
1
When "Where" Isn't Enough: Solving Real-World Problems with Behavioral Vectors
Dorothy Young Center for the Arts, Lancaster Rd, Madison, NJ, USPlease note that registration via EventBrite is required for in-person and on-line attendance.
We are happy to announce that Doug Drechsel will be our special guest at this month's meeting.
Abstract
Traditional location tracking systems are great at answering "where," but fail to answer "what is happening?" This creates a massive problem in real-world software engineering: operators are flooded with false alerts because a safe, normal event (like driving past a school) triggers the exact same alarm as a genuine threat (like loitering near a school).
In this talk, we’ll solve this context problem by building a practical, hands-on application of vector spaces. While vectors are incredibly popular right now for LLMs and text search, we will look at how to encode physical behavior - like speed, dwell time, and proximity - into a simple, 3-dimensional vector space.
You don't need any prior experience with spatial data to follow along. Using Oracle AI Database 26ai, we will walk through the database architecture and SQL required to combine basic location tracking with AI Vector Search. By the end of the session, you'll see how creating a simple vector space can reduce alert noise by up to 90% in high-stakes environments like parolee tracking and museum security.
What you will learn:
- Applied Vector Spaces: How to design and create a simple behavioral vector to solve a tangible logic problem
- The Context Gap: Why raw location data isn't enough, and how encoding behavior fixes it.
- Code in Action: A practical look at fusing spatial filters (SDO_GEOMETRY) and vector math (VECTOR_DISTANCE) in a single query to prioritize real threats without missing true positives.
Bio
Doug Drechsel has worked on many server side platforms including Tuxedo and WebLogic Server. He is currently a developer advocate working with the Database team and reaching out to full stack developers. His most recent work has been with creating servers as a service using Node.JS, specifically in the Mobile/Web use cases.
2 attendees
Past events
64
