On March 23-25, several members of Idibon’s technical staff attended the Association for the Advancement of Artificial Intelligence’s (AAAI) 2015 Spring Symposium. There are two symposia a year, one in the fall usually hosted in Arlington, Virginia and the other in the spring hosted by Stanford University’s Computer Science Department. The Symposia encourage meetings of researchers in smaller groups than the larger AAAI conference. The Symposia are also great testing grounds for new research areas, some of which use these workshop-style meetings as a jumping off point to eventually building a larger conference.
The Idibon team attended sessions in four of the eight symposia tracks. These represented a mix of our personal research interests and those that correspond with Idibon’s Human-in-the-Loop and Machine Learning based approach to text analytics. They were:
Structured Data for Humanitarian Technologies: Perfect Fit or Overkill?
Knowledge Representation and Reasoning: Integrating Symbolic and Neural Approaches
Socio-Technical Behavior Mining: From Data To Decisions
Turn-taking and Coordination in Human-Machine Interaction
While all four sessions were fantastic, the Structured Data for Humanitarian Technologies session presented an exciting opportunity to share our experiences working on disaster relief projects and the UNICEF U-report. Check out the session’s links and information on the use of structured data in humanitarian work.
Two of our favorite presentations covered the history of technology use in humanitarian projects and a novel approach to disaster monitoring based on Wikipedia article changes, respectively.
Sara Jayne Terp presented on Evolution of the Humanitarian Data Ecosystem (2004-2015)
Thomas Steiner presented on Natural Disaster Monitor based on real-time Wikipedia article updates. The source code and deployed demo are also available.
In addition to Sara Jayne and Thomas’s great presentations, we also discovered some cool resources that we wanted to share:
Humanitarian Data Exchange
DBPedia is “a crowd-sourced community effort to extract structured information from Wikipedia and make this information available on the Web. DBpedia allows you to ask sophisticated queries against Wikipedia, and to link the different data sets on the Web to Wikipedia data. We hope that this work will make it easier for the huge amount of information in Wikipedia to be used in some new interesting ways.”
Attending AAAI’s Spring Symposium was intellectually stimulating and socially great. We saw speakers from a variety of backgrounds – including the UN, Google, the US military – and learned more about how people with diverse perspectives come together to organize information during a crisis. At the same time, we revisited familiar algorithms (deep learning and neural nets) as well as new ideas, like Noah Goodman’s probabilistic programming language or ProPPR (a scalable way to apply first-order logic). It is important to us that we continue to evaluate our own technology against new paradigms.
Last month, we attended UC Berkeley Information School’s Infocamp, an unconference covering relevant topics such as encouraging data-intensive approaches to social good projects and improving Human Computer Interaction with thoughtful UI. We are looking forward to attending again next year.
In April, we will be attending and presenting at IAnnotate, Text By the Bay, and the Social Media and Web Analytics Innovation Summit. Please reach out to us if you are going to be attending these events!
– The Idibon Team
The post Humanitarian Technologies at AAAI – Presentations and Resources appeared first on Idibon.