Code from my project "Donde nacen las palabras: Análisis de neologismos y sus vecindarios semánticos a través de la semántica distribucional"
This study addresses the prediction of neology through distributional semantics using FastText, a computational model of distributional semantics. The main objective is to test Campbell's hypothesis in Spanish, which suggests that neologisms tend to emerge in areas of discourse with increasing popularity. Two areas of discourse were examined: technology and sciences (growing popularity) and philosophy and literature (stable popularity). The methodology involved extracting semantic neighborhoods from representative terms and identifying neologisms. The results indicate that more neologisms appear in the area of growing popularity than in the area of stable popularity. Although the difference is not statistically significant, the findings suggest a positive correlation between the growth in importance of a discourse area and the emergence of neologisms. The successful delineation of semantic neighborhoods strengthens the robustness of the methodological approach used, demonstrating the utility of distributional semantic models in the study of neology prediction. Despite limitations in sample size and other factors, this study contributes to the understanding of neology as a global and systematic phenomenon, focusing on extralinguistic factors (here, frequency growth). Future research could address these limitations and explore other explanatory factors of neology, as well as the automatic extraction of neologisms.
