Talia – first computational linguistics spin-off

Automatic language processing: TALIA, spin-off of the ‘Antonio Zampolli’ Institute of Computational Linguistics of the National Research Council (CNR) in Pisa in collaboration with GruppoMeta – ETT, a SCAI Group company. The new structure, the first in Italy in the sector, is a trait d’union between scientific research on Artificial Intelligence and the knowledge society

Donna illuminata dal riflesso di codici e stringhe

Pisa, 4 August 2022 – The TALIA spin-off created thanks to the activities of the ‘Antonio Zampolli’ Institute of Computational Linguistics of the CNR Pisa Department of Human and Social Sciences, Cultural Heritage (DSU) was presented in Pisa, in collaboration with GruppoMeta of ETT, a SCAI Group company.

The new structure, the first in Italy in this sector, is based on the conjugation of Artificial Intelligence (AI) with Automatic Language Processing (ALP), a technology leading an exponential growth, transversal to a multiplicity of knowledge society industries, from Public Administration to schools, healthcare and even the valorisation of intangible cultural heritage.

It is precisely the Automatic Language Processing systems, in fact, that, using Artificial Intelligence, transmit to a computer the indispensable skills for using natural language in order to interact with people (think chatbots). They also enable the automatic extraction of information from texts or other media (e.g. social media sentiment analysis ) and the dynamic extension of linguistic competence, as is the case with the instant translator.

Computational linguistics deals, in fact, with the analysis and processing of natural language, i.e. the language we usually use, through the use of computer methodologies. With automatic language processing systems, analyses can be made at speeds unimaginable for human capabilities.

Using the available tools, it is possible, for example, to understand what a text is about, how it talks about it and to identify the prevailing sentiment, the target audience, the mother tongue of the writer – even if it differs from that of the text – and, with some accuracy, even his or her age and level of schooling.

Simonetta Montemagnidirector of the ‘A. Zampolli Institute,’ recalls: “It is certainly no coincidence that TALIA was created within the ‘Antonio Zampolli’ Institute of Computational Linguistics of the CNR in Pisa, which has been active in the field of Computational Linguistics since 1967, as the Division of Computational Linguistics at the National University Centre for Electronic Computing (CNUCE), and since 1978 as an independent institute of the CNR. For years, it has played a key role in the development of technologies, products and services for natural language processing, with a focus on the Italian language. The Institute was among the first to undertake research in this field and, today, has an established position as a reference centre, both nationally and internationally’.

TALIA is the first computational linguistics spin-off that directly analyses and caters for the needs of the market, focusing on the experimentation and realisation of state-of-the-art models for the analysis of the Italian language (and not only) in real application contexts.

TALIA’s mission is, therefore, to achieve tangible growth and technological innovation in terms of functions, performance and quality in the areas of natural language processing and Artificial Intelligence techniques applied to texts.

Giovanni Verreschi, CEO of ETT Group, a SCAI Group company, which has controlled GruppoMeta since 2021, said: “Thanks to the specialisations of ETT and GruppoMeta in technology for innovation in the cultural sector, we have further verticalised our know-how in computational linguistics, making it possible for the Group to enter the world of digital solutions developed for content enhancement applied to cultural heritage and educational publishing. Specifically for this project, the market goal is to engineer NLP (Natural Language Processing) algorithms and technologies, the result of academic research, and make them easily usable through scalable and high-performance Cloud services in different sectors: from healthcare to publishing, from education to tourism, to public administration and, in general, for all those realities that need to catalogue and return a large amount of data‘.

Given the premises and expectations, the spin-off therefore aims to become the reference entity for all companies and realities that need to manage, store, analyse and catalogue large quantities of data and textual content – individual documents, manuals, databases, archives – in order to transform them into accessible information and/or to reorganise, understand and automatically rewrite them by exploiting the most modern information and Artificial Intelligence technologies. TALIA is also able to provide highly innovative services, both with respect to text analysis tools – in terms of content, sentiment, linguistic form – and in the authoring (for the realisation of a multimedia communication reproducible on a PC) and writing phases.

The aim of TALIA is precisely to bring together the results and expertise of all these years of research at the Institute of Computational Linguistics and exploit the state of the art in the field of Automatic Language Processing (ALP) and Artificial Intelligence (AI), to create software products and systems capable of solving any task that requires the analysis and production of text with the speed and reliability of a machine, but with the skills of a human being,” concludes Felice Dell’Orletta, president of Talia and first researcher at CNR.