Diarmuid Ó Séaghdha

Contact Details

Email: Consists of the strings "Diarmuid", "O'Seaghdha@cl", "cam", "ac" and "uk", all joined together by dots

Quick Links

Short Bio

UPDATE: I have moved to Apple, where I work on the Siri team. I remain affiliated with the Computer Lab as a Visiting Industrial Fellow and will continue to be involved in the NLIP group.

Previously I was a Senior Research Associate in the Natural Language and Information Processing Group at the Computer Laboratory in Cambridge. My interests revolve around the application of machine learning techniques to semantic processing tasks, especially tasks at the interface between lexical knowledge (what individual words mean) and relational knowledge (what combinations of words mean). Other areas of interest include linguistic conventionality, language use in social media and NLP applications in other sciences and the humanities.

In 2003 I graduated with a BA (Mod) in Computer Science, Linguistics and German from Trinity College Dublin. I subsequently spent a year studying Indian linguistic theory at the Faculty of Oriental Studies (now the Faculty of Asian and Middle Eastern Studies) in Cambridge under Dr. Eivind Kahrs. I then moved to the Computer Lab to do my PhD on Learning Noun Compound Semantics, which was supervised by Ann Copestake. I subsequently spent some time working with Anna Korhonen on an EPSRC-funded project "Lexical Acquisition for the Biomedical Domain" and with Simone Teufel on an IARPA-funded project "Foresight and Understanding from Scientific Exposition".

About my name (by popular request!)

My name is Irish in origin (like me). "Ó Séaghdha" is my surname; it is pronounced the same as "O'Shea", which is its anglicised form. Despite what my email address suggests, there is no apostrophe. A rough IPA transcription (based on the Wikipedia guidelines) of my full name is /dʲiərmˠʊɪdʲ oː ʃeː/. In BibTeX you need to write Diarmuid {\'O S\'eaghdha} or \'O S\'eaghdha, Diarmuid in the author field; otherwise BibTeX will assume incorrectly that the "Ó" is a middle initial and "Séaghdha" is my surname.

Compound noun bibliography

I have compiled a bibliography of computational and linguistic literature relating to compound nouns; it's quite out of date now but maybe someone will find it useful. There's also a Bibsonomy version which may be updated more frequently and has extras like indexible tags and abstracts.

Current and recent activities

We will be giving a tutorial on Learning Semantic Relations from Text at EMNLP 2015 in Lisbon. The slides are here.
Our paper Multi-domain Dialog State Tracking using Recurrent Neural Networks (with Nikola Mrkšić, Blaise Thomson, Milica Gašić, Pei-Hao Su, David Vandyke, Tsung-Hsien Wen and Steve Young) has been accepted at ACL 2015.
I will be giving a keynote talk on Compound interpretation as a problem for computational semantics at the 1st Workshop on Computational Approaches to Compound Analysis (co-located with COLING). My slides are here.
My paper Unsupervised learning of rhetorical structure with un-topic models (with Simone Teufel) has been accepted for oral presentation at COLING 2014.
I will be giving a talk on Computational Semantics for the Humanities at a workshop on Translation and the Digital: New Tools for Creativity and Communication at UCL on the 25th April. My slides are here.
Our paper Emoticons and Phrases: Status Symbols in Social Media (with Simo Tchokni and Daniele Quercia) has been accepted at ICWSM 2014.
My MPhil project proposals for 2013-14 are up. If you're interested in any of them, send me an email.
My paper Probabilistic Distributional Semantics with Latent Variable Models (with Anna Korhonen) has been accepted to appear in Computational Linguistics.
I (and some others) wrote a book! It's called Semantic Relations Between Nominals and it's about semantic relations between nominals.