A Network Approach For Interpreting Social Landscapes

点击翻译按钮获取中文摘要

Humans are in many ways defined by their social interactions with others. Indeed, the heart of human adaptive systems lies in peoples’ use of social networks to facilitate solutions to collective problems like resource shortfalls, information acquisition and dissemination, political turmoil, and conflict resolution. Material culture—clothing, artwork, jewelry—is often used to advertise important information about one’s personal and group identity within a social network. One way in which people do this is to adopt or manipulate the style of an object to distinguish themselves from others or more closely identify themselves with others. Archaeology is uniquely equipped, through the study of artifact style, to track the development of social networks across vast expanses of space and time. Within this broader context, Dr. Charles Egeland of the University of North Carolina at Greensboro, along with several colleagues, will develop a methodology based on artificial intelligence to identify stylistic similarities among digital images of artifacts. The level of stylistic similarity among artifacts will be input into custom-coded program plugins that can produce visual, graph-based representations of the pattern and strength of social ties between archaeological sites. Two graduate students will be trained to apply artificial intelligence and software integration in a social science setting. The data, artificial intelligence methodology, and plugins generated by this project will be freely accessible online by other users, and the technique can eventually be applied to any object represented by a 2D digital image. Ultimately, this project will help clarify how humans construct and use social networks to meet challenges. Dr. Egeland and his research team will examine how Paleolithic people used material culture to construct social networks and navigate the rapidly changing environments. The researcher will lead a team of archaeologists, paleoclimatologists, and computer scientists to (1) assemble a database of ~200 digital images of engraved artifacts; (2) construct an open-access, web-based application that uses artificial intelligence to identify stylistic patterns among the digital representations of the artifacts; and (3) develop custom plugins for an open-source graph analysis platform to produce visual representations of, and quantitative descriptors at multiple scales. With these data, the research team will explore how geography, environmental uncertainty, and population density influenced how people constructed and used their social networks to regulate social boundaries and offset resource shortages.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.