Customers: Main Archive of Moscow (Glavarchiv) Moscow; Government and social institutions Contractors: Yandex Product: Artificial intelligence (AI, Artificial intelligence, AI)Project date: 2022/06 - 2022/12
|
2022: Development of the "Search by Archives" platform
The capital's chief archive and Yandex have developed the Search by Archives platform. The neural network will be able to recognize handwritten texts in historical documents and translate them into digital format. The system will facilitate the search for citizens who study the history of a kind and actively use the online resource of the Glavarkhiv "My Family." The audience of this service for January 2023 is more than 10 million people. This became known on January 26, 2023.
Our joint project to introduce artificial intelligence into the field of archival businesses is needed to recognize and translate documents of the XVII-XIX centuries into a machine-readable format. The project was based on our online service "My Family" - it was on its materials that the developers trained the neural network. And if earlier the search took tens of hours, now the necessary names can be found in a matter of minutes. We hope that thanks to the service, the number of citizens interested in the history of their family will greatly increase. As of January 2023, 2.5 million pages of metric books and other genealogy documents are available on the Archive Search platform. In the future, their number will only grow, - said the head of the Moscow Glavarkhiv Yaroslav Onopenko. |
"Search Archives" will be an assistant in error-free reading of handwritten font. Metric documents were drawn up by hand, so researchers periodically have difficulties deciphering surnames and names, which can make further searches difficult. Now it is enough to drive names into the search string, and the system will show all the corresponding mentions. This will significantly save time on the compilation of the pedigree tree.
The main array of documents processed by the neural network was the materials of the capital's Glavarkhiv, but this service also included metric records from the archives of the Orenburg and Novgorod regions. According to the developers, over time, the number of vaults and available scanned files will increase.
The use of this technology will minimize the streaming of original documents, thereby protecting them from rapid decay. This will provide the city with the opportunity to preserve a documentary array about residents of Moscow and the Moscow province for future generations.
Prior to this, the search for data on the birth, marriage and death of citizens born before 1917 was carried out almost manually. The researcher needed to navigate the funds and affairs well, looking through large volumes of documents in the reading room of the Moscow Glavarchiv or in the online service "My Family," where more than eight million pages of metric books, revision tales and confession sheets are available to users.
The "Archive Search" service is not the first digital project implemented in the archive sphere. A few years ago, the virtual museum "Moscow - with concern for history" was opened, where you can see documents, objects and photographs transferred by residents for storage, as well as other archival materials from the funds of the capital's Glavarkhiv. In 2020, together with the State Inspectorate for Control over the Use of Real Estate in the city of Moscow, the project "Unique Documents" was created, which introduces documents about Moscow and its residents of great historical value. Digital developments in archival business create the most comfortable conditions and significantly save Muscovites time when collecting and obtaining the necessary information.