BGU, Microsoft embrace AI to advance Ben-Gurion archive research
To date, Microsoft has only applied its machine-learning solution to one other archive: thousands of documents relating to the assassination of President John F. Kennedy in 1963.
By EYTAN HALONDr. Adi Portughies presents an AI-based tool for the analysis of the Ben-Gurion Archive developed by Ben-Gurion University of the Negev and Microsoft Israel(photo credit: PR)
Prime minister David Ben-Gurion kept meticulous records of the events that shaped the formation and early days of the Jewish state.The diaries of the iconic founding father, some 20,000 pages of primary sources written over nearly seven decades, provide both a personal account of his life and an invaluable document for researchers of Zionism and Israeli history.In a pioneering move over 20 years ago, the Ben-Gurion Archive – at the Ben-Gurion Research Institute for the Study of Israel and Zionism at Ben-Gurion University of the Negev (BGU) – digitized the thousands of diary pages and other historical records. While accessible online to all, the sheer quantity of raw data has made comprehensive searching prohibitively difficult.But a groundbreaking partnership between BGU and Microsoft Israel launched two tools on Monday to “bring the archive to life.” The technologies are intended to provide greater insight to researchers and students alike.Together, the partners developed an advanced artificial intelligence-based system that will ultimately enable users to search the entire Ben-Gurion Archive, including both printed and handwritten materials, and construct causal relations, trace issues and ideas through multiple documents, and identify relationships between individuals and concepts through an interactive map.The tool, currently in its pilot stages, will enable previously impossible in-depth assessment of historical documents within minutes.The Ben-Gurion Archive and Microsoft Israel commenced the pilot program by applying its machine-learning solution to Ben-Gurion’s English-language correspondence with foreign leaders and members of the public following the Six Day War of June 1967. Hebrew-language sources, covering decades of Israeli and pre-state history, will be added in due course.“We want to share the knowledge and to develop new tools,” said Dr. Adi Portughies, head of the institute’s Infrastructure Information Systems department.“The main vision is to create an ecosystem where researchers, scholars, archivists and database administrators can share their databases. They can apply their tools and make a digital ecosystem for Israel studies.”A second tool integrates Ben-Gurion’s diary entries into Microsoft Outlook. Users can return to periods of interest in Israel’s history on Outlook’s calendar, and see the first Israeli premier’s diary entries according to the day and hour when they were written.The technologies were unveiled at an event launching Ben-Gurion University of the Negev and Microsoft Israel’s “Digital Humanities” program, advancing the adoption of new technologies in humanities research.“Microsoft Israel attributes great value to education and higher education, and therefore invests many resources in providing access to advanced technologies to universities, studies and researchers, such as artificial intelligence and the public cloud,” said Dr. Tomer Simon, National Technology Officer at Microsoft Israel.“As part of our collaboration with the Ben-Gurion Research Institute for the Study of Israel and Zionism, Microsoft Israel is assisting and accompanying the digital transformation process of the Ben-Gurion Archives to help create a broad research ecosystem.”To date, Microsoft has only applied its machine-learning solution to one other archive: thousands of documents relating to the assassination of President John F. Kennedy in 1963. Following an Act of Congress, the documents were declassified in 2017.The vast majority of the “JFK Files” collection – approximately 88% to date – has been released to the public, including documents withheld by the FBI, CIA and other agencies.
