Meaning induction is a classical and extremely relevant topic within various (cognitive) domains. While classical theories treated words as atomic units, recent frameworks provided support to the idea that also subword information can play a role, especially when a speaker encounters a word for the first time (i.e., novel words, that can be either existing or made up for experimental reasons). In the present article, after outlining some classic models of word meaning, we will describe more recent computational implementations (and related experimental evidence) that allow the representation of the semantic pattern elicited by strings of letters, regardless of their existence.
Gatti, D., Bonandrini, R. (2025). Subword Information in (Pseudo) Word Meaning. In Reference Module in Social Sciences. Elsevier.
Subword Information in (Pseudo) Word Meaning
Bonandrini, R
2025
Abstract
Meaning induction is a classical and extremely relevant topic within various (cognitive) domains. While classical theories treated words as atomic units, recent frameworks provided support to the idea that also subword information can play a role, especially when a speaker encounters a word for the first time (i.e., novel words, that can be either existing or made up for experimental reasons). In the present article, after outlining some classic models of word meaning, we will describe more recent computational implementations (and related experimental evidence) that allow the representation of the semantic pattern elicited by strings of letters, regardless of their existence.
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
