Aarne Talman, PhD

About me

Aarne

I am a language technology researcher and AI consultant with expertise in language understanding, reasoning, and large language models (LLM). Currently, work as a Data Science and Machine Learning Senior Manager at Accenture, where I leverage my 20 years of experience in research, software engineering, consulting, and leadership to drive technological advancements and to help organisations unlock the value of AI in real world applications.

I'm also a Visiting Scholar in Language Technology at the University of Helsinki. My research primarily centres around natural language understanding, reasoning, and natural language inference, employing machine learning techniques to address these challenges. I am particularly fascinated by the intricacies of language understanding, the development of AI models to represent it, and the methodologies for evaluating these models.

My educational background includes a PhD in Language Technology from University of Helsinki, an MSc in Computational Linguistics and Formal Grammar from King's College London, which I completed in 2007, and a BSc in Philosophy from the London School of Economics, obtained in 2005.

Contact me!

Research

My research focuses on natural language understanding (NLU), reasoning and natural language inference using machine learning. I’m interested in what understanding language consists of, how it can be modelled in AI and how these models should be evaluated. I have worked on a wide variety of different research topics in industry research labs during my career.

Natural Language Understanding

I have developed new NLU methods and models used in production by millions of end users. I have also studied the limitations of various benchmarks and datasets in NLU.

Speech Recognition and Prosody

I have conducted research in automatic speech recognition and prosody. I have developed speech recognition models used in production by millions of end users.

Neural Machine Translation

I have developed machine translation models and state-of-the-art neural machine translation systems used in production by millions of end users.

Publications

See my publications in: Google Scholar | ACL Anthology

Peer-Reviewed Papers

Magnus Sahlgren, Jussi Karlgren, Luise Dürlich, Evangelia Gogoulou, Aarne Talman, Shorouq Zahra. 2024. ELOQUENT 2024 — Robustness Task. Working Notes of the Conference and Labs of the Evaluation Forum (CLEF 2024). [bibtex][pdf]
Jussi Karlgren, Aarne Talman. 2024. ELOQUENT 2024 - Topical Quiz Task. Working Notes of the Conference and Labs of the Evaluation Forum (CLEF 2024). [bibtex][pdf]
Janek Bevendorf, Matti Wiegmann, Jussi Karlgren, Luise Dürlich, Evangelia Gogoulou, Aarne Talman, Efstathios Stamatatos, Martin Potthast, Benno Stein. 2024. Overview of the “Voight-Kampff” Generative AI Authorship Verification Task at PAN and ELOQUENT 2024. Working Notes of the Conference and Labs of the Evaluation Forum (CLEF 2024). [bibtex][pdf]
Jussi Karlgren, Luise Dürlich, Evangelia Gogoulou, Liane Guillou, Joakim Nivre, Magnus Sahlgren, Aarne Talman. 2024. ELOQUENT CLEF shared tasks for evaluation of generative language model quality. Advances in Information Retrieval. ECIR 2024. [bibtex]
Aarne Talman, Hande Celikkanat, Sami Virpioja, Markus Heinonen, Jörg Tiedemann. 2023. Uncertainty-Aware Natural Language Inference with Stochastic Weight Averaging. Proceedings of the 24th Nordic Conference on Computational Linguistics (NoDaLiDa). [bibtex] [pdf] [code]
Aarne Talman, Marianna Apidianaki, Stergios Chatzikyriakidis, Jörg Tiedemann. 2022. How Does Data Corruption Affect Natural Language Understanding Models? A Study on GLUE datasets. Proceedings of the 11th Joint Conference on Lexical and Computational Semantics. [bibtex] [pdf] [data and code]
Aarne Talman, Marianna Apidianaki, Stergios Chatzikyriakidis, Jörg Tiedemann. 2021. NLI Data Sanity Check: Assessing the Effect of Data Corruption on Model Performance. Proceedings of the 23rd Nordic Conference on Computational Linguistics. [bibtex] [pdf] [data and code]
Aarne Talman, Antti Suni, Hande Celikkanat, Sofoklis Kakouros, Jörg Tiedemann and Martti Vainio. 2019. Predicting Prosodic Prominence from Text with Pre-trained Contextualized Word Representations. Proceedings of the 22nd Nordic Conference on Computational Linguistics. [bibtex] [pdf] [corpus and code]
Aarne Talman, Umut Sulubacak, Raúl Vázquez, Yves Scherrer, Sami Virpioja, Alessandro Raganato, Arvi Hurskainen, and Jörg Tiedemann. 2019. The University of Helsinki submissions to the WMT19 news translation task. Proceedings of the Fourth Conference on Machine Translation: Shared Task Papers. [bibtex] [pdf]
Aarne Talman and Stergios Chatzikyriakidis. 2019. Testing the Generalization Power of Neural Network Models Across NLI Benchmarks. Proceedings of the 2019 ACL Workshop BlackboxNLP: Analyzing and Interpreting Neural Networks for NLP. [bibtex] [pdf]
Aarne Talman, Anssi Yli-Jyrä and Jörg Tiedemann. 2019. Sentence Embeddings in NLI with Iterative Refinement Encoders. Natural Language Engineering 25(4). [bibtex] [pdf] [code]

Preprints

Risto Luukkonen, Jonathan Burdge, Elaine Zosa, Aarne Talman, Ville Komulainen, Väinö Hatanpää, Peter Sarlin, Sampo Pyysalo. 2024. Poro 34B and the Blessing of Multilinguality. arXiv. [bibtex] [pdf] [model and code]

Theses

Aarne Talman. 2024. Towards Natural Language Understanding: Developing and Assessing Approaches and Benchmarks. Doctoral Dissertation. University of Helsinki.
Aarne Talman. 2006. Path Grammars and the Generative Capacity of Dynamic Syntax. Master's Dissertation. King's College London.
Aarne Talman. 2005. A Limit on Artificial Intelligence? - The Gödelian Case. Bachelor's Thesis. The London School fo Economics and Political Science.