exaly project was started by a few researchers to address the absence of a comprehensive scholarly search engine. There are several services, but none of them provides all the services researchers may need. Most of them are commercial products and require subscriptions for using all features.
We started this project to address our needs, but we then decided to make this service accessible to the entire research community. It took a few years to build the necessary infrastructure to extend our home project to the production level.
exaly is a non-profit project and commercially disinterested. We never used any public fund or commercial support by corporations at any stage of the project. This is indeed a scientific project. As described on the licence page, all materials are available under a Creative Commons licence.
Along with the search capabilities, we compute primary scientometric measures.
Our ultimate goal is to provide all the services researchers need in one place and make them freely available to everyone.
How does it work?
We use various sources to double-check the accuracy of data. Whenever the full text of an article is not available online, the PDF files are directly parsed to index the text and extract the references. Although we check with all available resources, we parse each and every reference for capturing the citations.
The exaly project has been chiefly developed in C from scratch. No external library has been used whatsoever. MariaDB was utilised as the database management system. For natural processing language (NLP) tasks, we used Stanford NLP only for double-checking our own system.