Seeking Meaning in a Space Made out of Strokes, Radicals, Characters and Compounds (1104.4321v1)

Abstract: Chinese characters can be compared to a molecular structure: a character is analogous to a molecule, radicals are like atoms, calligraphic strokes correspond to elementary particles, and when characters form compounds, they are like molecular structures. In chemistry the conjunction of all of these structural levels produces what we perceive as matter. In language, the conjunction of strokes, radicals, characters, and compounds produces meaning. But when does meaning arise? We all know that radicals are, in some sense, the basic semantic components of Chinese script, but what about strokes? Considering the fact that many characters are made by adding individual strokes to (combinations of) radicals, we can legitimately ask the question whether strokes carry meaning, or not. In this talk I will present my project of extending traditional NLP techniques to radicals and strokes, aiming to obtain a deeper understanding of the way ideographic languages model the world.