Free

Constructing an academic knowledge graph and recommendation engine

Event Information

Share this event

Date and Time

Location

Location

Postgraduate Hub training room 2

First Floor

Senate House

University of Bristol

BS8 1TH

United Kingdom

View Map

Event description

Description

Given only freely available text, we can extract sufficient data to create knowledge graphs, representing both individual components and collectives as a whole. These graphs can be used to identify key ideas, overlapping concepts and areas of missing information. From here they can recommend specific events, identify communities, derive values for missing data and predict areas of change over time. The fundamentals of this approach, however, lie in extracting reliable and representative data for each individual component.

One specific example of this can be found in academic publications. By using unique user IDs and public data we can construct a knowledge graph for a defined set of people, e.g. a department or University.

What to expect

This workshop will demonstrate the techniques and methods required to do this, including the use of APIs, extracting and enriching informative text, natural language processing and constructing recommendation engines. We will show how this kind of approach can be used to recommend collaborations, automatically identify people matching a specific piece of text and identify topic areas with high and low coverage.

Data Week 2019

This event is part of Data Week 2019, run by the Jean Golding Institute

Any questions? Please contact jgi-coordinator@bristol.ac.uk

Share with friends

Date and Time

Location

Postgraduate Hub training room 2

First Floor

Senate House

University of Bristol

BS8 1TH

United Kingdom

View Map

Save This Event

Event Saved