Navigation ☰ About Contact Research Interests CV Publications

Philip John Gorinski, looking stern. In January 2019, I started a position as Research Scientist at Huawei Noah's Ark Lab in London, working on Natural Language Processing.

Before joining the industry, I was a PhD student at the Institute of Language, Logic and Cognition (ILCC), University of Edinburgh, working on Automatic Movie Analysis and Summarization. My supervisors were Mirella Lapata and Rik Sarkar.

After successfully completing my PhD, I continued working in postdoc positions at the University of Edinburgh, first with Mirella Lapata on further multi-modal film/TV analysis, later with Beatrice Alex on Entity Recognition for Medical Health Records.


You can send me e-mails. I'm usually quick to reply: p [dot] j [dot] gorinski [at] gmail [dot] com

Visit my LinkedIn profile. It is almost up to date.

I am also on Xing. Might be less up to date.

And Research Gate. Definitely not up to date.

I always plan to upload code to GitHub, it just does not happen.

You can come see me on Facebook!

Click here for my Twitter. 0 < tweets < 100 in 10 or so years, but I do respond there.

I'm also on Mastodon (, I guess. Expected activity even less than on Twitter.
Research Interests

My general areas of research are Natural Language Processing and Machine Learning. I use computational models applied to natural language (text, audio) as well as visual resources (image, video).

Some of my current work includes research into Reinforcement Learning for Code Synthesis LLMs, as well as projects on parameter-efficient finetuning.

In my first postdoc position I employed deep learning methods to the discovery and modelling of plot structure and events in TV crime shows such as CSI - Crime Scene Investigation.
My second postdoc was part of a joint project between the University of Edinburgh and NHS Scotland, developing text mining methods for medical health records related to brain imaging, for which we developed specialised Named Entity Recognition Networks.
For my PhD thesis, I developed dynamic computational models to generate extractive summaries for movie scripts, as well as visual summaries of feature films. I also employed deep learning to the tasks of analysing the content of movies as well as generating short, informative movie overviews.

In the olden days, before starting my PhD and the takeover of NLP by Neural Networks and Deep Learning, my main interests were in grammar formalisms, parsing, and Computational Semantics. My MSc thesis dealt with machine translation using semantic mappings in Head-Driven Phrase Structure Grammar (HPSG), for my BSc I focussed on structure mapping via bimorphisms.


Below you can have a look at my short-form curriculum vitae. My full CV is available as PDF.

05/2024 - ongoing NLP Research Engineer, Robin AI, London
01/2019 - 05/20204 Senior Research Scientist NLP, Huawei Noah's Ark Lab, London
01/2018 - 01/2019 Postdoc Researcher, University of Edinburgh, School of Informatics, Institute for Language, Cognition and Computation
09/2013 - 11/2017 PhD Candidate, University of Edinburgh, School of Informatics, Institute for Language, Cognition and Computation
06/2015 - 09/2015 Research Intern, Microsoft Research, Redmond, WA. Research in the Natural Language Processing group. Supervisors: Chris Quirk, Michel Galley.
03/2011 - 08/2013 M.Sc. Program, Saarland University, Sarbruecken, Germany. M.Sc. Program "Language Science and Technology". Final grade: 1.4 (excellent).
04/2013 - 09/2013 Student Freelancer, DIaLOGIKa Software GmbH, Saarbrücken. Software Consulting and Testing.
2010 - 2012 Student Research Assistant, SemEval, Cluster of Excellence, Saarland University. Research on NullInstantiations of FrameNet frames, programming work.
09/2006 - 03/2011 B.Sc. Program, Saarland University, Saarbrücken, Germany. B.Sc. Program "Computational Linguistics". Final grade: 1.5 (excellent).
2009 - 2010 Student Research Assistant, Project “IDIX”, Cluster of Excellence, Saarland University. Research on idioms in context, programming work and annotation.
Winter 2008/09 B.Sc. Program, University of Edinburgh, Edinburgh, Scotland. ERASMUS study period abroad as part of B.Sc. program.
2007 - 2008 Student Research Assistant, Project “SALSA” Department of Computational Linguistics, Saarland University. Semantic annotation of german newspaper texts, using a german version of FrameNet.
1998 - 2006 Abitur, Paul-Klee Gymnasium, Overath, Germany.
(Click on titles to see abstracts)

Mojtaba Valizadeh, Philip John Gorinski, Ignacio Iacobacci, Martin Berger (2024). Correct and Optimal: the Regular Expression Inference Challenge. IJCAI 2024, Jeju, South Korea. [PDF]

Philip John Gorinski, Matthieu Zimmer, Gerasimos Lampouras, Derrick Goh Xin Deik, Ignacio Iacobacci (2023). Automatic Unit Test Data Generation and Actor-Critic Reinforcement Learning for Code Synthesis. Findings of EMNLP, Singapore. [PDF]

Yunjie He, Philip John Gorinski, Ieva Staliūnaitė, Pontus Stenetorp (2023). Graph Attention with Hierarchies for Multi-hop Question Answering. Preprint, arXiv. [PDF]

Alexander I. Cowen-Rivers, Philip John Gorinski, Aivar Sootla, Asif Khan, Liu Furui, Jun Wang, Jan Peters, Haitham Bou Ammar (2022). Structured Q-learning For Antibody Design. RL4RealLife Workshop at NeurIPS2022, New Orleans, LA, USA. [PDF]

Ieva Staliūnaitė, Philip John Gorinski, Ignacio Iacobacci (2022). Relational Graph Convolutional Neural Networks for Multihop Reasoning: A Comparative Study. Preprint, arXiv. [PDF]

Ieva Staliūnaitė, Philip John Gorinski, Ignacio Iacobacci (2021). Improving Commonsense Causal Reasoning by Adversarial Training and Data Augmentation. AAAI2021, Virtual. [PDF]

Yusheng Tian, Philip John Gorinski (2020). Improving End-to-End Speech-to-Intent Classification with Reptile. InterSpeech2020, Shanghai, China. [PDF]

Rafael Kourdis, Gabriel Gordon-Hall, Philip John Gorinski (2020). αVIL: Learning to Leverage Auxiliary Tasks for Multitask Learning. Preprint, arXiv. [PDF]

Gabriel Gordon-Hall, Philip John Gorinski, Shay B. Cohen (2020). Learning Dialog Policies from Weak Demonstrations. Proceedings of ACL 2020, Seattle, WA, USA. [PDF]

Gabriel Gordon-Hall, Philip John Gorinski, Gerasimos Lampouras, Ignacio Iacobacci (2020). Show Us the Way: Learning to Manage Dialog from Demonstrations. The Eighth Dialog System Technology Challenge (DSTC8) Workshop at AAAI2020, New York, NY, USA. [PDF]

Philip John Gorinski, Honghan Wu, Claire Grover, Richard Tobin, Conn Talbot, Heather Whalley, Cathie Sudlow, William Whiteley, and Beatrice Alex (2019). Named Entity Recognition for Electronic Health Records: A Comparison of Rule-based and Machine Learning Approaches. Presented at Healthcare Text Analytical Conference (HealTAC) 2019, Cardiff, Wales. [PDF]

Philip John Gorinski (2018). Automatic Movie Analysis and Summarization. PhD Thesis at the University of Edinburgh, School of Informatics, ILCC, Edinburgh, Scotland. [PDF]

Philip John Gorinski and Mirella Lapata (2018). What's this Movie about? A Joint Neural Network Architecture for Movie Content Analysis. In Proceedings of NAACL-HLT 2018, New Orleans, LA, USA. [PDF]

Philip John Gorinski and Mirella Lapata (2015). Movie Script Summarization as Graph-based Scene Extraction. In Proceedings of NAACL-HLT 2015, Denver, CO, USA. [PDF]

Philip John Gorinski, Josef Ruppenhofer, and Caroline Sporleder (2013). Towards Weakly Supervised Resolution of Null Instantiations. In Proceedings of IWCS 2013 - Long Papers, Potsdam, Germany. [PDF]

Vera Demberg, Asad B. Sayeed, Philip J. Gorinski, and Nikolaos Engonopoulos (2012). Syntactic surprisal affects spoken word duration in conversational contexts. In Proceedings of EMNLP-CoNLL '12, Jeju Island, South Korea. [PDF]

Josef Ruppenhofer, Philip John Gorinski, and Caroline Sporleder (2011). In search of missing arguments: A linguistic approach. In Proceedings of of RANLP 2011, Hissar, Bulgaria. [PDF]

Caroline Sporleder, Linlin Li, Philip John Gorinski, and Xaver Koch (2010). Idioms in Context: The IDIX Corpus. In Proceedings of LREC 2010, Valletta, Malta. [PDF]

This website aims to be HTML5/CSS4 compliant and has been validated using the W3 nu validator.
If you find any glaring issues with code or scripts, please let me know so I can fix it :)