Evaluating the semantic representation contextualization of polysemic words

BIANCHI BRUNO; LAURINO JULIETA; KACZER LAURA; KAMIENKOWSKI JUAN

San Luis

Congreso; Congreso de la Sociedad Argentina de Neurociencias; 2023

Sociedad Argentina de Neurociencias

Polysemy (words with the same spelling and different, but related, meaning) is a featurepresent in many languages. For example, in English it is estimated that 80% of words arepolysemic. This generates that, when we come across a word, either visually or aurally,our brain must quickly interpret what meaning is being referred to, based on the contextin which it appeared. The mechanisms that operate to carry out this processing are notprecisely known.Meanwhile, state-of-the-art computational Language Models have achieved a level oflanguage comprehension that allows them to recognize the correct meaning that isbeing assigned to polysemic words when used in context.In the present work we present a preliminary comparative analysis between behavioralresponses of humans and the mechanisms that operate in Language Models against theinterpretation of neutral sentences (which do not allow defining the meaning of thepolysemic words present) contextualized with biasing paragraphs. This work is a firststep towards a better understanding of the brain mechanisms underlying this process