Ricerc@Sapienza

Contextualized word embeddings have been employed effectively across several tasks in Natural Language Processing, as they have proved to carry useful semantic information. However, it is still hard to link them to structured sources of knowledge. In this paper we present ARES (context-AwaRe Embeddings of Senses), a semi-supervised approach to producing sense embeddings for the lexical meanings within a lexical knowledge base that lie in a space that is comparable to that of contextualized word vectors.

The knowledge acquisition bottleneck problem dramatically hampers the creation of sense-annotated data for Word Sense Disambiguation (WSD). Sense-annotated data are scarce for English and almost absent for other languages. This limits the range of action of deep-learning approaches, which today are at the base of any NLP task and are hungry for data.

Accurate semantic representation models are essential in text mining applications. For a successful application of the text mining process, the text representation adopted must keep the interesting patterns to be discovered. Although competitive results for automatic text classification may be achieved with traditional bag of words, such representation model cannot provide satisfactory classification performances on hard settings where richer text representations are required.

Word senses are typically defined with textual definitions for human consumption and, in computational lexicons, put in context via lexical-semantic relations such as synonymy, antonymy, hypernymy, etc. In this paper we embrace a radically different paradigm that provides a slot-filler structure, called “semagram”, to define the meaning of words in terms of their prototypical semantic information. We propose a semagram-based knowledge model composed of 26 semantic relationships which integrates features from a range of different sources, such as computational lexicons and property norms.

Notes on semiotics and ethology: a (partially interrrupted diagolgue. The success of the chomskyan paradigm, combined with the anthropocentric turn of semiotics, has determined, since the seventies, a sort of divorce between semiotic studies and the ethological research. However, ethology, especially in its variant known as ‘cognitive’ ethology, has continued to use semiotic terminology in its approach to the communicative behavior of non-human animals.

The paper aims at discussing some cases of metaphorical linguistic expressions in some ancient Indo-European languages, in order to show how conceptual metaphors (in the sense of Lakoff and Johnson 1980) “materialize” in certain linguistic expressions through the use of concepts such as spatiality and embodiment.

The aim of this paper is to analyze the semantics of the Old Persian verb duruj-, usually translated as ‘to lie’, to deceive’. In order to better understand the semantic nuances and the contextual meanings of the Old Persian verb and to offer a comprehensive – albeit concise – description, this paper is divided into two main sections. The first part is devoted to a brief overview of various cultural-historical aspects starting from Iranian mythology, which – as is known – is closely linked to the world of ritual.

This paper is an essay in micro-historical semantics and concerns the Late Latin verb charaxāre, a loanword from Gr. χαράξαι, χαράσσω. The scientific literature uses to separate two different semantic values of this Latin verb: ‘to carve; to cut’ and ‘to write’. In Herren’s opinion, starting from Virgilius Maro Grammaticus’ writings, charaxāre ‘to write’ was unique to Hisperic usage and derived from the knowledge of Greek glosses circulating among Irish monks from the VI-VII century AD onwards.