The paper focuses on methods of segmentation and information extraction from natural language texts for automated retrieval of economy needs in new scientific, technical and technological solutions. We provide an analytical overview of modern machine learning based methods and outline a framework for economy needs retrieval. The proposed framework combines bootstrapping with manual hypotheses confirmation and active learning.
Download Bulletin of Russian Academy of Natural Sciences № 3 / 2017, page 3 (in Russian): https://raen.info/upload/redactorfiles/maket_vestnik_2017_03.indd%20(1).pdf
Burak P., Zvorykina T., Kormalev D., Zhebel V. Information extraction methods for the automated retrieval of economy needs in new scientific, technical and technological solutions // Bulletin of the Russian Academy of Natural Sciences. 2017.Vol. 17. No. 3. P. 3-8.