A novel chatbot architecture forthe Italian language is presented that is aimed at implementing cognitive under-standing of the query by locating its cor-respondent subgraph in the agent’s KB bymeans of a graph matching strategy pur-posely devised. The FCG engine is usedfor producing replies starting from the se-mantic poles extracted from the candidateanswers’ subgraphs. The system imple-ments a suitable disambiguation strategyfor selecting the correct answer by analyz-ing the commonsense knowledge relatedto the adverbs in the query that is embed-ded in the lexical constructions of the ad-verbs themselves as a proper set of fea-tures. The whole system is presented, anda complete example is reported throughoutthe paper.
This article is authored also by Synbrain data scientists and collaborators. READ THE FULL ARTICLE