$$$$
{{ $t($store.state.user.experience_value_in_dollars) }}
Expert
{{ $t($store.state.user.experience_search_name) }}
0
jobs
Data scientist with experience of Natural Language Processing
Peter Bleackley
,
Horsham, United Kingdom
Experience
Other titles
Skills
I'm offering
I have undertaken data science and machine learning projects for clients ranging from startups to multinationals. I have a particular interest in Natural Language Processing - for one client I created a Computational Linguistics pipeline that achieved state of the art accuracy for Word Sense Disambiguation. I have particularly strong experience in research and development projects.
Markets
United Kingdom
Links for more
Once you have created a company account and a job, you can access the profiles links.
Industries
Language
English
Fluently
Available
My experience
2020 - 2021
freelance
Data Scientist
Sopra Steria.
Working with HSBC to develop a software platform to improve the efficiency of model training for anti-money-laundering systems.
Python, Jupyter, Pandas, Data Science
2020 - 2020
freelance
Data Scientist
Amey.
Worked on a project to develop data science applications for Highways England. Led the development of an automatic diagnostic system for traffic flow sensors, using Isolation Forests
Machine learning, Python, Artificial Intelligence
2019 - 2019
freelance
NoSQL Developer
Cambridge Quantum Computing Ltd.
(Playful Technology Limited)
Developing all aspects of the MongoDB database for a quantum random number
generation system, including data capture and postprocessing. Specified protocol
for transfer of data between capture hardware and database, and implemented in C++. Investigated and implemented randomness extraction algorithm. Enabled
rapid capture and processing of large volumes of data.
Developing all aspects of the MongoDB database for a quantum random number
generation system, including data capture and postprocessing. Specified protocol
for transfer of data between capture hardware and database, and implemented in C++. Investigated and implemented randomness extraction algorithm. Enabled
rapid capture and processing of large volumes of data.
MongoDB, Database, C, NoSQL, Technology, Developer, Hardware, Processing
2018 - 2019
freelance
Consultant
Pentland Brands.
Investigating machine learning methods for recommending the optimum style of swimming goggles given a 3D model of a face and user metadata. Created a test framework into which various data reduction algorithms and machine learning classifiers could be placed and results compared. Data reduction and classification algorithms were mainly based on scikit-learn, however a graph convolutional
network was implemented in Theano. After finding no significant correlations between facial shapes and goggle preferences, I recommended the work be discontinued.
network was implemented in Theano. After finding no significant correlations between facial shapes and goggle preferences, I recommended the work be discontinued.
Machine learning, Python, Artificial Intelligence
2018 - 2018
freelance
Data Scientist
Rolls Royce AI Hub.
Developed a testbed to compare topic modelling algorithms for a proposed search system. During this work I discovered a bug in the Gensim topic modelling library and contributed a fix. Worked on a parser library in Python to extract structured data from scanned and machine readable PDF documents. Advised on the technical aspects of external tenders (including sitting on the tender panel) and a proposed
collaborative project.
collaborative project.
Python, Data Science, Data mining
2018 - 2018
freelance
Independent Consultant
All Street Research.
Using Natural Language Processing techniques to mine corporate documents for paragraphs related to key themes. Used Gensim, NLTK, Pandas, Numpy, Keras and Jupyter Notebooks. This was a short term project, undertaken while awaiting security clearance for the role at Rolls Royce.
Natural language processing, Artificial Intelligence, Machine learning, Python, Data Science
2017 - 2017
freelance
Contract Data Scientist
Boehringer Ingelheim.
Using Natural Language Processing techniques to mine online forums fordata related to health conditions and pharmacovigilance. Used MongoDB and Gensim.
Natural language processing, Python, Data Science, Machine learning, Artificial Intelligence
2017 - 2017
freelance
Contract Data Scientist / NLP Specialist
True 212.
Developing a computational linguistics pipeline as part of a Real-time
Editorial Resource system. Developed machine learning components for named entity recognition, part of speech tagging, word sense disambiguation and topic modelling. Word sense disambiguation achieved 70% accuracy, which is considered state-of-the-art performance. This will be used to identify related content during the creation of articles for the client's news website, and predict
engagement. Used MongoDB, Numpy, Scipy,Scikit-learn, Pandas, Hidden Markov Models, and Gensim. Work funded by the Google Digital News Initiative.
Editorial Resource system. Developed machine learning components for named entity recognition, part of speech tagging, word sense disambiguation and topic modelling. Word sense disambiguation achieved 70% accuracy, which is considered state-of-the-art performance. This will be used to identify related content during the creation of articles for the client's news website, and predict
engagement. Used MongoDB, Numpy, Scipy,Scikit-learn, Pandas, Hidden Markov Models, and Gensim. Work funded by the Google Digital News Initiative.
Natural language processing, Python, Data Science, Machine learning, Artificial Intelligence
2017 - 2017
freelance
Contract Software Developer
Techbit.
Developing algorithms to classify swimming strokes and extract performance metrics from motion sensor data. This was a short-term project, as the client was tendering the system to a third party, and did not win funding for further development.
Machine learning, Artificial Intelligence, Python, Software development
2016 - 2017
freelance
Contract Data Scientist
Formisimo.
Developing machine learning algorithms to predict conversion of online forms, using Hidden Markov Models, Support Vector Machines and LSTM networks (Keras). The client had previously found that models that appeared promising in initial tests would fail in more realistic simulations. I was able to explain this by investigating the data in more depth and devise models that performed better by taking this insight into account. Since early results were promising, the project
was extended from 3 months to 4 1⁄2.
was extended from 3 months to 4 1⁄2.
Data Science, Python, Machine learning, Artificial Intelligence
2016 - 2016
freelance
Contract Data Scientist
Social Finance.
Data exploration, restructuring and data cleansing of a public sector datasetmusing Pandas and Jupyter Notebooks.
Python, Data Science
2016 - 2016
freelance
Computational Linguistics Contractor
Metafused.
Research and Development for a system to extract semantic data from free text. Natural language processing, machine learning, supervised learning.
Natural language processing, Python, Data Science, Machine learning, Artificial Intelligence, MongoDB
2015 - 2015
freelance
Contract Data Scientist
Valtech.
Provided consultancy services to a music-based social network startup.
Developed a naive Bayes classifier that determined whether or not metadata from different music streaming services referred to the same tune.
Developed a naive Bayes classifier that determined whether or not metadata from different music streaming services referred to the same tune.
Python, Data Science, Machine learning, Artificial Intelligence
2014 - 2015
job
Mathematical Software Developer
Arithmetica Limited.
Algorithm development for lidar analysis and vector model fitting software.
C++ development for a Windows application, and rapid prototyping in Python.
C++ development for a Windows application, and rapid prototyping in Python.
Python, Prototyping, C, Rapid Prototyping, Windows, Developer, Development, Software
2013 - 2014
job
Product Innovation and Experience Lead
HumanLearning Limited.
Developing content and behaviour based recommendation for a social
learning /business communication product, using Python.
learning /business communication product, using Python.
Python, Innovation, Content, Social
2013 - 2013
freelance
Contractor
Zooey Consulting.
Investigated specialist document analysis technology.
Technology
2011 - 2013
job
Chief Data Analyst
MulTplx Limited.
Developed Python software that infers customer behaviour from interactions with display devices in mobile phone shops, and visualisations (HTML 5 and Javascript) and analyses of the data gathered. Clients included Samsung and EE.
Javascript, Html, Python, HTML/CSS/Javascript, Analyst, Software
2001 - 2011
job
Research Engineer
BBC Research and Development.
Projects included:
• Extracting semantic data from free text. I created a database of timed
segments of programmes from the BBC's archive catalogue, and investigated techniques for analysing subtitles, and applications for this data.
I later reimplemented one of the algorithms I had investigated and contributed
it to the Gensim topic modelling library.
Used Python, HTML, XML and JSON.
• A European Collaborative project to investigate Affective Computing
techniques for interactive storytelling. http://callas-newmedia.eu
• Artificial intelligence techniques for broadcast chain diagnostics. Identified
suitable technology, and prototyped, using C++.
• Investigating suitable parameters for a low-delay MPEG 2 coder. Software
simulation to inform hardware design using C++.
• Hardware implementation of the Dirac video codec. Created Open Source
circuits for arithmetic coding / decoding and exp-golomb coding /decoding.
Refined algorithms to make them suitable for hardware implementation.
• Extracting semantic data from free text. I created a database of timed
segments of programmes from the BBC's archive catalogue, and investigated techniques for analysing subtitles, and applications for this data.
I later reimplemented one of the algorithms I had investigated and contributed
it to the Gensim topic modelling library.
Used Python, HTML, XML and JSON.
• A European Collaborative project to investigate Affective Computing
techniques for interactive storytelling. http://callas-newmedia.eu
• Artificial intelligence techniques for broadcast chain diagnostics. Identified
suitable technology, and prototyped, using C++.
• Investigating suitable parameters for a low-delay MPEG 2 coder. Software
simulation to inform hardware design using C++.
• Hardware implementation of the Dirac video codec. Created Open Source
circuits for arithmetic coding / decoding and exp-golomb coding /decoding.
Refined algorithms to make them suitable for hardware implementation.
Database, Http, Software, Hardware, Implementation, It, Open source, Technology, Algorithms, C, Html, Storytelling, Artificial Intelligence, XML, JSON, Research, Video, HTML/CSS/Javascript, Python, Design
My education
?
-
2000
University of Leicester
Doctorate, Astrophysics
Doctorate, Astrophysics
?
-
1996
University of Durham
N/a, M.Sci. Physics
N/a, M.Sci. Physics
?
-
1992
Thornleigh College
Secondary, N/a
Secondary, N/a
Peter's reviews
Peter has not received any reviews on Worksome.
Contact Peter Bleackley
Worksome removes the expensive intermediaries and gives you direct contact with relevant talent.
Create a login and get the opportunity to write to Peter directly in Worksome.
38000+ qualified freelancers
are ready to help you
Tell us what you need help with
and get specific bids from skilled talent in Denmark