Nestor Alvaro (Data Scientist at Deloitte Digital, Spain)
Applying dependency parsing to extract relations
between elements in the sentences

Date: May 17, 2019 (Friday)
Time: 15:00 - 17:00 (Japan Time)
Venue: UTokyo, Komaba Campus, Building 2, Room 303
Language: English

There are different levels of informations that can be extracted from the written information. Starting with a Part-Of-Speech (POS) processing, to identify of the nouns, verbs and other syntactic elements, we then moved to Named Entity Recognition (NER), to determine which of these were referring to people, locations and organization. We moved further to ground the locations to determine the mentioned city and country. At this point we had a good amount of information, but we still have to go one step further to really understand who is performing which action, and where.

In this workshop we present a tool that can be used to extract all these elements: POS tags, NER elements and also the Dependency Parse. This tool will be further use in following steps in the project.

Nestor Alvaro is a Data Scientist at Deloitte Digital researching on machine learning (ML) and natural language processing (NLP).

*Supported by JSPS KAKENHI, Project/Area Number 15KT0040  (Takeshi Wada, The University of Tokyo)