텍스트마이닝을 통해 분석한 지역별 고유 구비설화 전승의 면모 -지명전설과 인물전설을 중심으로

Abstract

This study identified and cataloged the legends of place names and historical figures from a compilation of oral narratives collected from nine regions nationwide, as documented in the 『Comprehensive Korean Oral Literature』 and the 『Complementary Edition of Comprehensive Korean Oral Literature』. For this, text mining techniques were employed, involving the analysis process of “data collection → preprocessing of regional information → preprocessing of proper nouns → analysis of regional folklore → data generation.” First, we collected data on 26,542 oral narrative titles and preprocessed the regional information, which was not organized by administrative divisions at the “provincial” level. Subsequently, we categorized the data into nine regions. Next, we conducted a task to identify proper nouns that were not recognized by the morphological analyzer and registered them as nouns. In this process, we used the “associated word network” analysis technique to identify proper nouns. Once the preprocessing task of verifying and modifying the proper nouns was completed, we extracted the frequency of the nouns in each region. Then, to select nouns that guarantee the uniqueness of the region, the frequency was “normalized” and the value was “scaled.” Additionally, the criterion for selecting the unique place names and people of each region was defined by the term “maximum regional share” and selected as 80%.
 As a result, 520 nouns were extracted, of which only proper nouns corresponding to place and personal names were filtered. Based on this, for legends of place names, place names were classified into “natural features,” “artificial features,” and “village names” for each region, according to the “nature of the evidence.” This list has been compiled. In the legends of historical figures, the corresponding personal names are presented for each region. Legends of place names are crucial as they reveal how residents of each region shape and give meaning to their lives in the “space where history intersects.” However, legends of historical figures deserve attention for their exploration of individuals in the region who have been marginalized or excluded from historical awareness accumulated over time and documented in written records.