User Tools

Site Tools


openings

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revision Previous revision
openings [2016/11/01 17:23]
gilles
openings [2016/11/09 15:06] (current)
gilles [Objectives]
Line 15: Line 15:
  
  
-The PhD student will work on the **Data Analysis and Semantic Extraction** work package, within the Knowledge engineering team ([[http://​cui.unige.ch/​isi|KE@ISS]]) of the University of Geneva'​s Center for computing ([[http://​cui.unige.ch|CUI]])+The PhD student will work on the **Data Analysis and Semantic Extraction** work package, within the Knowledge engineering team ([[http://​cui.unige.ch/​isi]]) of the University of Geneva'​s Center for computing ([[http://​cui.unige.ch]])
  
-This step consists in applying information extraction (IE) techniques to attach semantics to the analysed raw data coming from IoT objects, and to produce semantic descriptors (attribute-value pairs) that are needed by the semantic firewall. The analysed raw data are (scanned) images, digital sound recording, measurements (collected by sensors), etc., accompanied by first level analysis results (document layout, component tags, extracted texts, ...) Each category of raw data may in turn represent different types of content. For instance, an IoT scanned document may be a national ID card, a payslip, a medical record, etc. The aim of this WP is to devise an information extraction technique that 1) recognizes the type of content represented in the raw data and 2) adapts to the content type and utilizes background knowledge (eg. an ontology of content types) to efficiently extract the information that characterize it’s content (specific attribute values such id card number, account number, product name, ...) It will be based on a combination of techniques from the fields of information extraction and knowledge representation.+This work consists in applying information extraction (IE) techniques to attach semantics to the analysed raw data coming from IoT objects, and to produce semantic descriptors (attribute-value pairs) that are needed by the semantic firewall. The analysed raw data are (scanned) images, digital sound recording, measurements (collected by sensors), etc., accompanied by first level analysis results (document layout, component tags, extracted texts, ...) Each category of raw data may in turn represent different types of content. For instance, an IoT scanned document may be a national ID card, a payslip, a medical record, etc. The aim of this work package ​is to devise an information extraction technique that 1) recognizes the type of content represented in the raw data and 2) adapts to the content type and utilizes background knowledge (eg. an ontology of content types) to efficiently extract the information that characterize it’s content (specific attribute values such id card number, account number, product name, ...) It will be based on a combination of techniques from the fields of information extraction and knowledge representation.
  
 ==== Application ==== ==== Application ====
openings.1478017416.txt.gz · Last modified: 2016/11/01 17:23 by gilles