• A
  • A
  • A
  • АБВ
  • АБВ
  • АБВ
  • А
  • А
  • А
  • А
  • А
Обычная версия сайта

Мероприятия

Cеминар ММИТ: доклады Liubov Tupikina и Tristan Miller

Мероприятие завершено

Liubov Tupikina выступит с докладом "Flow-networks: graph theoretical approach to study flow systems"
Tristan Miller выступит с докладом "Natural language processing with UIMA and DKPro"

Title: Flow-networks: graph theoretical approach to study flow systems
Speaker: Liubov Tupikina, Ecole Polytechnique (Paris, France)
Time:16:40-18:10
Abstract: Complex network theory provides an elegant and powerful framework to statistically investigate different types of systems such as society, brain or the structure of local and long-range dynamical interrelationships in the climate system. Network links in correlation, so-called climate networks typically imply information, mass or energy exchange. However, the specific connection between oceanic or atmospheric flows and the climate network’s structure is still unclear. We propose a theoretical approach of flow-networks for verifying relations between the correlation matrix and the flow structure, generalizing previous studies and overcoming the restriction to stationary flows [1]. We studied a complex interrelation between the velocity field and the correlation network measures. Our methods are developed for correlations of a scalar quantity (temperature, for example) which satisfies an advection-diffusion dynamics in the presence of forcing and dissipation. Our approach reveals the insensitivity of correlation networks to steady sources and sinks and the profound impact of the signal decay rate on the network topology. We illustrate our results with calculations of degree and clustering for a meandering flow resembling a geophysical ocean jet. Moreover, we discuss the follow-up approaches and application of the flow-networks method [2].

[1] "Correlation networks from flows. The case of forced and time-dependent advectiondiffusion dynamics" L.Tupikina, N.Molkenthin, C.Lopez, E.Hernandes-Garcia, N.Marwan, J.Kurths, Plos One. 2016

[2] "A geometric perspective on spatially embedded networks. Quantification of edge anisotropy and application to flow networks", H.Kutza, N.Molkenthin, L.Tupikina, J.Donges, N.Marwan, U.Feudel, J.Kurths, R.Donner, Chaos, 2016

Title: Natural language processing with UIMA and DKPro
Speaker:
Tristan Miller, Technische Universität Darmstadt (Germany)
Time:18:10-19:40

Abstract: This talk introduces UIMA (Unstructured Information Management Architecture), an industry-standard software architecture for content analytics.  UIMA provides extensible data, component, and process models for annotating, exchanging, and analyzing unstructured data such as natural-language text.  We also introduce DKPro, a family of ready-to-use natural language processing (NLP) components built on UIMA. Using UIMA and DKPro, students and researchers can rapidly develop and deploy experimental text processing pipelines.  In a classroom setting, these tools are valuable because they significantly reduce the barriers to entry for learning and applying advanced NLP techniques.  Using DKPro, students can start projects in text classification, discourse analysis, etc., without needing to spend time implementing lower-level NLP tasks such as morphological analysis, word sense disambiguation, or text similarity.  In a graduate-level research setting, UIMA and DKPro facilitate conducting experiments in a fully reproducible manner.  The talk will provide a tutorial-style overview of both frameworks, including code snippets and sample applications.

Место проведения: Кочновский проезд, 3. ауд.317

Приглашаются все желающие. Для студентов, аспирантов, преподавателей и сотрудников Высшей школы экономики - вход свободный.

При необходимости заказа пропуска в здание НИУ ВШЭ просьба сообщить по e-mail: lantropova@hse.ru.