Daniel Blackwell

Key Skills

Software

AWS SageMaker

Top Subject Matter

Textual customer feedback classification

Web scraped online reviews

Survey data

Top Data Types

Text

Top Task Types

Classification

Freelancer Overview

An experienced senior data scientist who has worked on all aspects of the machine learning project lifecycle, from labelling to model evaluation. At a previous company, I built a labelling tool in AWS SageMaker to speed up classifying text information: whereby unlabelled text was ordered by confidence of being part of the class being labelled, meaning the items least like those being labelled before, were labelled first.

Labeling Experience

Text labelling for NLP classifiers

Aws SagemakerTextClassificationEmotion Recognition

I labelled tens of thousands of customer feedback examples into multiple categories, to be able to build NLP classifier models. Most comments needed to be classified into multiple categories, so to speed this up, I built a widget tool in AWS. Behind the widget a model ordered unlabelled text, every 50 classifications, so we could label the text that a model would be least confident in categorising.

I labelled tens of thousands of customer feedback examples into multiple categories, to be able to build NLP classifier models. Most comments needed to be classified into multiple categories, so to speed this up, I built a widget tool in AWS. Behind the widget a model ordered unlabelled text, every 50 classifications, so we could label the text that a model would be least confident in categorising.

2018 - 2018

Education

Work History