Translating the Talmud is being incredibly challenging. This text is like a vast deposit of human knowledge; being aware of its importance and strength helped us facing such a valuable challenge.
That’s why we created the multi-disciplinary task force of professionals working as a whole on this massive translation project, supported by highly-advanced computational linguistics tools.

The Italian Jewish community accepted the challenge and it’s proving to have the outstanding professional skills required to engage such a challenging work. So far, only the largest and best-established Jewish communities such as the American, Israeli, French, Russian and German took on similar tasks. This unique idea of using a smart software, developed specifically for this project, to translate the Talmud is being successful in every respect. And it’s completely Italian.

Right from the very start, the Italian State realized the great importance of this project and through its authorities gave a significant contribution by providing highly skilled professionals and the resources needed.
The commitment of the Italian State to the Talmud translation project legitimates the Italian scientific community and the universal value of the Talmud, and represents a unique gift for the Nation, its citizens and for all those who love and speak our language across the globe.

The Talmud originates from recording in writings the discussions of academics debating the Torah in Babylon; it became popular in the following three centuries, then eventually was subjected to studies, commentaries and analysis within all Jewish communities across the globe, until present days.

The Talmud translation project can be described as extremely complex and sophisticated, the result of the Talmudic culture matching computational linguistics: two different forms of knowledge held together by the Talmud, true foundation of the Jewish identity.

Therefore, the project reveals its universal value of key to access a source of knowledge that so far was not available in Italian.

The CNR’s Institute for Computational Linguistics developed the web application TRADUCO specifically for the Babylonian Talmud Translation Project. This system is designed to process the text and knowledge according to the principles of computational linguistics; also, it includes highly advanced editorial tools.

The system TRADUCO is:

• accessible via web: contrary to desktop applications, TRADUCO simply needs Internet connection and can be used through a navigation browser;
• teamwork friendly: this web-based, highly technological application allows the team of translators, proofreaders and editors to work on the same data simultaneously and share them with each others;
• intelligent: the application includes a support system suggesting automatically the most appropriate translation for each sentence, thus delivering enhanced time optimization and homogeneous final result;
• flexible: this application is provided with a probability-based system of text analysis and processing adapting to different languages. Also, TRADUCO can easily be modified and reused to translate any kind of texts.

In addition, TRADUCO allows to:

• insert notes, commentaries and quotes;
• highlight the text on a semantic level, according to specific categories (such as people, rabbis, nature, measures, etc.);
• carry out complex researches on the brief text;
• read layouts created ad hoc and export translated texts for immediate production of paper copies.

TRADUCO, like any other system developed by a research institute, is being constantly upgraded. The suggestion component is the core of this innovative translation system. TRADUCO is already able to automatically supply the translator with the right translation, on the base of the archive of text that has already been translated by the translator him/herself or by other translators: after three years and a half of work, its memory is now enhanced to an extent that the system is able to suggest the correct translation, one out of two times. The system can be integrated with further information levels enhanced by an evaluation process capable of improving the translation accuracy and “rating” the quality of the suggestions provided (stored in the so-called “translation memory”). Therefore, the translation memory can be enhanced with: I) bilingual dictionaries, to translate in Italian (and set once and for all) words or idioms of the different linguistic evolutions of the Hebraic-Aramaic; II) algorithms using linguistic and semantic information to improve the process of recovery of the translations to suggest; III) a software component implementing the methodology of the evaluation adopted.

Also, the ILC team skills in knowledge engineering will be useful to integrate the System with semantic-based systems for text research and browsing. Indeed, by organizing the Talmud key words in a formal structure, the base of Talmudic knowledge, encoded by computer experts, can be created (for example, through the ontology formalism) to enable the access to the text contents to answer complex questions such as for example: “where in the text do botanical terms related to respiratory diseases appear?”, “where in the text do botanical terms related to the preparation of specific seasonal food or to specific events and feast days appear?”, “which are the most relevant topics debated in the commentaries to certain passages of the Talmud discussed by a specific rabbi in a specific period?”

Scientific Coordinator: Emiliano Giovannetti
Technical Coordinator: Davide Albanesi
Research and Development: Andrea Bellandi – Alessia Bellusci – Giulia Benotto – Francesco Bulleri – Enrico Carniani