NE:Tech - February 2020

February brings NE:Tech to us the day before Valentine's day, the day for celebrating love, and if you have a love for tech then NE:Tech is the place to be.

Please welcome your speakers:

Stephen Powers, Software Engineer @ SAGE

• Limitations of Sentiment Analysis
• Basic experience required
• Tags: Natural Language Processing

Sentiment Analysis has become an increasingly popular tool for businesses, public bodies and researchers in aiding decision making by mining data for useful information. With a rush of natural language processing service providers entering the market, we must ask ourselves ‘how accurate can sentiment analysis really be?’.

This talk sets out to explore this very question by looking at data containing political opinions, which exposes one of sentiment analysis’s greatest weaknesses – sarcasm. We will discuss briefly the methodology of building a simple API to collect tweets, the algorithm used in this study to analyse sentiment and the key findings.

A little about Stephen:
Previously worked in the public and political sector with a strong interest in driving the digital economy, particularly here in the North East. Went back to University to complete a master’s in computer science completing a thesis focused on Sentiment Analysis of data containing political opinion. Since finishing University in July 2019, has worked at SAGE as a developer.