Last week we attended the OSD FOCUS 2010 Human Social Cultural Behavior (HSCB) Modeling Program Conference in Chantilly, VA. It was an oversubscribed event with more than 400 attendees reflecting the broad interest in this program. Plenary presentations stressed the need to focus on the needs of the warfighter as opposed to the needs of the research community. The importance of data especially at the province level was drilled over and over as as a true need. Several speakers urged the adoption of common meta-data for social science data sets. Another related modeling issue addressed was the difference between correlation (coincidence) vs. causal relationships. Joe Watts of AGC lamented the lack of GIS support in the HSCB projects, and urged the development of HSCB map symbology for transition to the warfighter. Dr. Lisa Costa described the use of POET (Political, Operational, Economic and Technical) relationships in social network analysis, and the importance of developing a social radar for HSCB analysis. In his concluding remarks, Dr. Robert Foster challenged everyone to think about solutions for the professional training of HSCB domain.
We participated in a poster exhibition and gave three presentations: first, a general overview of our Semantic Wiki for Complex Operations project; second, the knowledge management needs of the complex operations community; and, third, open source social science data repository for HSCB research. Our presentations were received well and generated several questions. In particular, a number of attendees expressed the suitability of our wiki supporting the training of complex operations professionals. There was also an interest in our social data repository supporting legacy data and automating the ingestion of new data.
There were too many interesting presentations to cover every one. Dr. Sean O'Brien gave an overview of his Conflict Modeling, Planning, and Outcomes Experimentation (COMPOEX) program. The precision and recall performance of the lead contractor Lockheed Martin's system based on Ada boosting of multiple social behavior models was fairly impressive. Dr. Barry Silverman's agent based model CountrySim - one of the models in the Lockheed Martin system - was one of the models contributing to the high performance. In terms of data collection, Prof. Mansoor Moaddel's values and attitudes survey for Afghanistan, Iraq, Iran, Saudi Arabia, Jordan and Egypt was particularly interesting.
CDR Dylan Schmorrow put the state of the art in human social culture behavior models into perspective by comparing the HSCB models to models for weather and economic forecasting. The maturity of weather forecasting models is higher than those for economic forecasting. On this scale, CDR Schmorrow positioned HSCB modeling on the x-axis while hoping the HSCB program serving as a catalyst to ignite HSCB modeling into a steady march towards maturity.