Our paper relates the application of Natural Language Processing and Data Science to interrogate ethnography based, in part, on innovations from Kinship Theory developed by Dwight Read and colleagues.
iKLEWS (Infrastructure for Knowledge Linkages from Ethnography of World Societies) is a HRAF project funded by the National Science Foundation. iKLEWS developed semantic infrastructure and associated services for a growing textual database (eHRAF World Cultures) of ethnography. Semantic and data mining infrastructure developed under iKLEWS assists in identifying diverse, universal and cross-cultural aspects of a wide range of user selected topics, such as social emotion and empathy, economics, politics, use of space and time, morality, or music and songs.
Critically, iKLEWS is concerned with understanding ethnography as ethnographic theory, identifying data, knowledge and circumstances underlying the wide array of means by which ethnographers use textual exposition to relate what they understood to be important information about people.
In this paper we relate how the principles underlying the generation and representation of kinship terminologies, as developed by Dwight Read and his colleagues over the past 40 years, can be applied to pan-corpus semantic relations as produced by numerous ethnographers in numerous ethnographies. For kinship terminologies this approach has demonstrated that terminologies can be represented algebraically based on a subset of the terminology, which can then generate the entire terminology from the algebra derived from this partial accounting.
We are applying this approach to analysis of semantic graphs, ranging from sentential level dependency graphs to pan-corpus hypergraphs of semantically related terms that reflect the composite product of ethnographer judgements on language use in describing various topics. While we will utilise these structures in examining the depiction of kinship related instances in the text, we are interested in broader patterns of semantic relations, and comparing differences in these between ethnographers.