|
This app is released as part of a research project described by Khishigsuren, Regier, Vylomova and Kemp (2025), A computational analysis of lexical elaboration across languages.
The app allows you to explore lexical elaboration scores for 6000 concepts across 616 languages. Our analyses use a standard version of the BILA dataset that includes 8,575 lemmatized nouns, and the most frequent 6000 are included in this app. The left menu allows you to choose a concept. The chart below that menu shows languages with the top Llang scores for that concept, and languages with scores in the top 5% for that concept are shown as red points on the map below the chart. Terms with distributions over dictionaries similar to that of the chosen concept are shown at the bottom of the left column. Some interesting cases to explore include snow, wind, smell, dance, mountain, canoe, sheep, kangaroo, beer, coffee, porridge, ghost, angel, and god. The right menu allows you to choose a language. The chart below that menu shows concepts with the greatest Llang scores for that language, and the map below that chart shows the location of that language. Languages with distributions over concepts similar to that of the chosen language are shown at the bottom of the right column. Most analyses in our paper use a set of words associated with a concept of interest: for example, the results in Figure 1a are based on combined frequencies for snow and snowfall. This app, however, only shows results for single words, which means that results generated using the app will be similar but not always identical to results included in the paper. The words associated with some languages may reflect views that are harmful and offensive. The app is based on data from many dictionaries, some of which were published more than two centuries ago, and the results should not be taken to endorse the views expressed by any of these dictionaries. |