About the Center

In the past few years, the world has witnessed the breakthrough of generative language models with impressive AI capabilities, such as ChatGPT and others. These models represent the culmination of over ten years of research into NLP (natural language processing) and deep learning.

At Afeka, researchers have long been dedicated to exploring AI and NLP, focusing on written and spoken language in both English and Hebrew. Established more than a decade ago, the Afeka Center for Language Processing (ACLP) conducts applied research in language and voice processing and is recognized as a leader in the field, both in Israel and internationally. Building on ACLP's extensive expertise, the center now adds a new dimension: integrating large language models (LLMs) with voice processing technologies. It serves as a vital hub of knowledge and resources for researchers and industries working in NLP, offering: As part of the center's work, efforts are focused on researching and developing automated tools to enhance existing text repositories and generate tailored datasets. This includes creating a toolkit with customized text processing models and developing innovative metrics to assess the quality of outputs generated by language models.

In addition to its research activities, the center offers consulting services to industry and defense organizations, engages in community outreach, and supports nonprofit initiatives.

The Center’s Activity

  • Adapting and training existing Hebrew language models

    Adapting and training existing Hebrew language models

  • Developing advanced architectures for purpose-built language processing models (e.g., adding logic to the legal field, a pedagogic methodology to teaching, etc.)

    Developing advanced architectures for purpose-built language processin…

  • Developing specific datasets using automated tools

    Developing specific datasets using automated tools

  • Research and development of new tools and metrics

    Research and development of new tools and metrics

  • Integrating speech with generative language models such as GPT

    Integrating speech with generative language models such as GPT

  • Testing the sensitivities and limitations of existing models while developing new metrics for evaluating the quality of results

    Testing the sensitivities and limitations of existing models while dev…

Center Director

Dr. Sharon Yalov-Handzel

Dr. Sharon Yalov-Handzel

Dr. Yalov-Handzel brings 30 years of experience in AI, spanning diverse R&D roles in high-tech industries and academia. She earned her graduate degree from the Weizmann Institute and her doctorate from Bar-Ilan University, where her research focused on robotics and AI. Her professional work in high tech centered on robotics, with an emphasis on developing algorithms and solving mathematical models for robot motion planning, as well as navigation and calibration systems for autonomous systems with varying kinematic structures. In recent years, Dr. Yalov-Handzel has served as a consultant to high-tech companies and organizations in data science and NLP. Alongside her consulting work, she continues to contribute to academia as a lecturer and researcher specializing in advanced AI topics.

Center Members