Stergios Chatzikyriakidis

Stergios Chatzikyriakidis
University of Crete | UOC · Department of Philology

BA MSc PhD Docent

About

103
Publications
29,228
Reads
How we measure 'reads'
A 'read' is counted each time someone views a publication summary (such as the title, abstract, and list of authors), clicks on a figure, or views or downloads the full-text. Learn more
597
Citations
Introduction
Stergios Chatzikyriakidis currently is a Professor of Computational Linguistics at the University of Crete
Additional affiliations
December 2019 - present
University of Gothenburg
Position
  • Managing Director
April 2016 - November 2017
University of Gothenburg
Position
  • Project Manager
April 2016 - November 2019
University of Gothenburg
Position
  • Project Manager
Education
September 2006 - September 2010
King's College London
Field of study
  • Theoretical Linguistics
September 2005 - September 2006
King's College London
Field of study
  • Computational Linguistics and Formal Grammar
September 2000 - March 2005
Aristotle University of Thessaloniki
Field of study
  • Classics and Linguistics

Publications

Publications (103)
Article
Full-text available
In this paper we propose a way to deal with Natural Language Inference (NLI) by implementing Modern Type Theoretical semantics in the proof-assistant Coq. The paper is a first attempt to deal with NLI and Natural Language reasoning in general by using proof-assistant technology. Valid NLIs are treated as theorems and as such the adequacy of our acc...
Conference Paper
Full-text available
We discuss the semantics of NL coordination in modern type theories (MTTs) with coercive subtyping. The issue of conjoinable types is handled by means of a type universe of linguistic types. We discuss quantifier coordination, arguing that they should be allowed in principle and that the semantic infelicity of some cases of quantifier coordination...
Article
Full-text available
In this paper, a full account of clitic positioning in Cypriot Greek is attempted within the framework of Dynamic Syntax. Firstly, it is shown that the existing approaches dealing with CG clitic positioning are inadequate to deal with the full range of clitic positioning phenomena as these are described by Pappas (2010) and Chatzikyriakidis (2010)....
Article
Full-text available
Conversation is full of non-sentential utterances, with seamless shifts between speaker/hearer roles. The hurdle these pose for standard assumptions is that every linguistic dependency can be distributed across speakers, the content of what they say emerging incrementally. We argue that their modelling necessitates recasting language in terms of ou...
Conference Paper
Full-text available
In this paper we revisit the issue of copredica-tion from the perspective of modern type theories. Specifically, we look at: a) the counting properties of dot-types, and b) the case of a complex dot-type that has remained unsolved in the literature, i.e. that of newspaper. As regards a), we show that the account proposed in (Luo, 2010) for dot-type...
Conference Paper
Full-text available
In this paper, we present a number of finegrained resources for Natural Language Inference (NLI). In particular, we present a number of resources and validation methods for Greek NLI and a resource for precise NLI. First, we extend the Greek version of the FraCaS test suite to include examples where the inference is directly linked to the syntactic...
Conference Paper
Full-text available
The usage of social media platforms has resulted in the proliferation of work on Arabic Natural Language Processing (ANLP), including the development of resources. There is also an increased interest in processing Arabic dialects and a number of models and algorithms have been utilized for the purpose of Dialectal Arabic Natural Language Processing...
Chapter
In this paper, we look at the issue of gradability within formal semantics in modern type theories (MTT-semantics). Specifically, we look at both gradable adjectives and nouns, and show that the rich typing mechanisms afforded by MTT-semantics can give us a natural account of gradability. Gradable adjectives take indexed nouns as their arguments, w...
Preprint
Full-text available
A central question in natural language understanding (NLU) research is whether high performance demonstrates the models' strong reasoning capabilities. We present an extensive series of controlled experiments where pre-trained language models are exposed to data that have undergone specific corruption transformations. The transformations involve re...
Article
Full-text available
In this paper, the historical development of the clitic systems of Standard Modern, Cypriot and Pontic Greek are discussed. These three varieties not only present the whole range of variation one can find across clitic systems in Greek but, furthermore, derive from a common linguistic ancestor, i.e. Koine Greek. This paper argues that the transitio...
Conference Paper
Full-text available
In this paper, we propose an implementation of temporal semantics that translates syntax trees to logical formulas, suitable for consumption by the Coq proof assistant. The analysis supports a wide range of phenomena including: temporal references, temporal adverbs, aspec-tual classes and progressives. The new semantics are built on top of a previo...
Conference Paper
Full-text available
Formal semantics in the Montagovian tradition provides precise meaning characterisa-tions, but usually without a formal theory of the pragmatics of contextual parameters and their sensitivity to background knowledge. Meanwhile, formal pragmatic theories make explicit predictions about meaning in context, but generally without a well-defined compo-s...
Preprint
Full-text available
Pre-trained neural language models give high performance on natural language inference (NLI) tasks. But whether they actually understand the meaning of the processed sequences remains unclear. We propose a new diagnostics test suite which allows to assess whether a dataset constitutes a good testbed for evaluating the models' meaning understanding...
Article
Full-text available
In this paper, we propose a framework for dealing with anaphora and ellipsis which is both general and algorithmic. We ensure this generality by interpreting dynamic semantics phenomena as side effects, using a monad structure. We additionally implement this framework as an algorithm which translates abstract syntax to logical formulas. We argue th...
Preprint
Full-text available
In this paper, we propose an implementation of temporal semantics which is suitable for inference problems. This implementation translates syntax trees to logical formulas, suitable for consumption by the Coq proof assistant. We support several phenomena including: temporal references, temporal adverbs, aspectual classes and progressives. We apply...
Article
In everyday conversation, no notion of “complete sentence” is required for syntactic licensing. However, so-called “fragmentary”, “incomplete”, and abandoned utterances are problematic for standard formalisms. When contextualised, such data show that (a) non-sentential utterances are adequate to underpin agent coordination, while (b) all linguistic...
Conference Paper
Full-text available
In this paper, we explore the idea that independently developed Dynamic Syntax accounts of dialogue and interaction fit well within the general approach of radical embodied and enac-tive accounts of cognition (REEC). This approach enables a rethinking of the grounding of linguistic universal constraints, specifically tree structure restrictions, as...
Conference Paper
Full-text available
In this paper, I look at the case of afterthoughts and claim that these can receive a natural syntactic explanation once they are seen as clarification strategies, more specifically, as answers to implicit clarification questions. I argue that this assumption paired with a framework that is suited to deal with structures that go beyond sentential s...
Conference Paper
Full-text available
Natural Language Inference models have reached almost human-level performance but their generalisation capabilities have not been yet fully characterized. In particular, sensitivity to small changes in the data is a current area of investigation. In this paper, we focus on the effect of punctuation on such models. Our findings can be broadly summar...
Conference Paper
Full-text available
As the number of social media users increases, they express their thoughts, needs, socialise and publish their opinions. For good social media sentiment analysis, good quality resources are needed, and the lack of these resources is particularly evident for languages other than English, in particular Arabic. The available Arabic resources lack eith...
Conference Paper
Full-text available
In this paper, we propose a method to modify natural textual entailment problem datasets so that they better reflect a more precise notion of entailment. We apply this method to a subset of the Recognizing Textual Entailment datasets. We thus obtain a new corpus of entailment problems, which has the following three characteristics: 1. it is precise...
Chapter
Full-text available
In this paper we investigate the use of Deep Learning (DL) methods for Dialectal Arabic Sentiment Analysis. We propose a DL model that combines long-short term memory (LSTM) with convolutional neural networks (CNN). The proposed model performs better than the two baselines. More specifically, the model achieves an accuracy between 81% and 93% for b...
Conference Paper
Full-text available
We present the Shami-Senti corpus, the first Levantine corpus for Sentiment Analysis (SA), and investigate the usage of off-the-shelf models that have been built for Modern Standard Arabic (MSA) on this corpus of Dialectal Ara-bic (DA). We apply the models on DA data, showing that their accuracy does not exceed 60%. We then proceed to build our own...
Conference Paper
Full-text available
We present a system for Natural Language Inference which uses a dynamic semantics converter from abstract syntax trees to Coq types. It combines the fine-grainedness of a dynamic semantics system with the pow-erfulness of a state-of-the-art proof assistant. We evaluate the system on all sections of the FraCaS test suite, excluding section 6. This i...
Conference Paper
Full-text available
Neural network models have been very successful in natural language inference, with the best models reaching 90% accuracy in some benchmarks. However, the success of these models turns out to be largely benchmark specific. We show that models trained on a natural language inference dataset drawn from one benchmark fail to perform well in others, ev...
Conference Paper
Full-text available
We present BIS, a Bayesian Inference Semantics , for probabilistic reasoning in natural language. The current system is based on the framework of Bernardy et al. (2018), but departs from it in important respects. BIS makes use of Bayesian learning for inferring a hypothesis from premises. This involves estimating the probability of the hypothesis,...
Conference Paper
Full-text available
In this paper, we look at Natural Language Inference, arguing that the notion of inference the current NLP systems are learning is much narrower compared to the range of inference patterns found in human reasoning. We take a look at the history and the nature of creating datasets for NLI. We discuss the datasets that are mainly used today for the r...
Article
Full-text available
Copredication, especially when combined with quantification, provides interesting examples to support the idea that common nouns have their own identity criteria, as once argued for by Geach and subsequently studied by others. In this paper, revisiting the use of dot-types in modern type theories to model copredication, we show that, when both copr...
Preprint
Full-text available
In this paper, we present a new corpus of entailment problems. This corpus combines the following characteristics: 1. it is precise (does not leave out implicit hypotheses) 2. it is based on "real-world" texts (i.e. most of the premises were written for purposes other than testing textual entailment). 3. its size is 150. The corpus was constructed...
Preprint
Full-text available
In this paper, we present a new corpus of entailment problems. This corpus combines the following characteristics: 1. it is precise (does not leave out implicit hypotheses) 2. it is based on "real-world" texts (i.e. most of the premises were written for purposes other than testing textual entailment). 3. its size is 150. The corpus was constructed...
Article
Full-text available
Diglossia is a very common phenomenon in Arabic-speaking communities, where the spoken language is different from both Classical Arabic (CA) and Modern Standard Arabic (MSA). The spoken language is characterised as a number of dialects used in everyday communication as well as informal writing. In this paper, we highlight the lexical relation betw...
Preprint
Full-text available
Neural network models have been very successful for natural language inference, with the best models reaching 90% accuracy in some benchmarks. However, the success of these models turns out to be largely benchmark specific. We show that models trained on natural language inference dataset drawn from one benchmark fail to perform well in others, eve...
Presentation
Full-text available
Invited talk at the Language Technology group, University of Helsinki
Conference Paper
Full-text available
We propose a compositional Bayesian semantics that interprets declarative sentences in a natural language by assigning them probability conditions. These are conditional probabilities that estimate the likelihood that a competent speaker would endorse an assertion, given certain hypotheses. Our semantics is implemented in a functional programming l...
Presentation
Full-text available
Part of the exploration course "New type-theoretic tools in natural language semantics" at the North American Summer School on Logic, Language, and Information, CMU 2018.
Conference Paper
Full-text available
In type-theoretical semantics, sentences may often be interpreted as judgements, rather than propositions. When interpreting composite sentences such as those involving negations and conditionals, one may want to turn a judgemental interpretation into a proposition in order to obtain an intended semantics. In this paper, we propose a new negation o...
Conference Paper
Full-text available
Modern Standard Arabic (MSA) is the official language used in education and media across the Arab world both in writing and formal speech. However, in daily communication several dialects depending on the country, region as well as other social factors, are used. With the emergence of social media, the dialectal amount of data on the Internet have...
Presentation
Three day Tutorial at the Centre of Advanced Study, Oslo. The material for the tutorial can be found here: https://github.com/StergiosCha/CoqNL
Article
Full-text available
In this paper, we show how a rich lexico-semantic network which has been built using serious games, JeuxDeMots, can help us in grounding our semantic ontologies in doing formal semantics using rich or mod- ern type theories (type theories within the tradition of Martin Löf). We discuss the issue of base types, adjectival and verbal types, hyperonym...
Research
Full-text available
Abstract for the Workshop on Approaches to Coercion and Polysemy. Oslo, 2017.
Conference Paper
Full-text available
We present a type-theoretical framework for formal semantics, which leverages two existing well established tools: Grammatical Framework (GF) and Coq. The framework is the semantic equivalent of GF's resource grammars: every syntactic construction is mapped to a (compositional) semantics. Our tool thus extends the standard GF grammar with a formal...
Article
Full-text available
In this paper, we discuss a number of issues in adjectival and adverbial modification from the perspective of Modern Type Theories (MTTs). Specifically, our aim is to show how modification can be approached within a formal semantics language that assumes common nouns to be types rather than predicates, involves elaborate typing structures (e.g. dep...
Conference Paper
Full-text available
We explore prerequisites necessary for embedding Dynamic Syntax within an account of language evolution. We show how the dynamics of processing as modelled in Dynamic Syntax display remarkable parallelism with Clark's (2016) Pre-dictive Processing Model and that the interactive stance of a combined DS/PPM model of language/cognition reflects the Mu...
Conference Paper
Full-text available
In this talk, modification (both adjectival and adverbial) in Dynamic Syntax (Kempson et al. 2001, Cann et al. 2005) is going to be discussed. The literature on modification in DS is rather poor and no proper discussion has ever been attempted so far besides sparse mentions and incomplete in passing accounts. For example, in Cann (2002), adjectives...
Article
Full-text available
In this paper we present a study of adjectival/adverbial modification using modern type theories (MTTs), i.e. type theories within the tradition of Martin-Löf. We present an account of various issues concerning adjectival/adverbial modification and argue that MTTs can be used as an adequate language for interpreting NL semantics. MTTs are not only...
Chapter
Full-text available
When type theories are used for formal semantics, different approaches to the interpretation of common nouns (CNs) become available w.r.t whether a CN is interpreted as a predicate or a type. In this paper, we shall first summarise and analyse several approaches as found in the literature and then study a particularly interesting and potentially ch...
Book
Full-text available
This book is a collective volume that reports the state of the art in the applications of type theory to linguistic semantics. The volume fills a 20 year gap from the last published book on the issue and aspires to bring researchers closer to cutting edge alternatives in formal semantics research. It consists of unpublished work by some key researc...
Conference Paper
Full-text available
In this paper we discuss the use of interactive theorem provers (also called proof assistants) in the study of natural language semantics. It is shown that these provide useful platforms for NL semantics and reasoning on the one hand, and allow experiments to be performed on various frameworks and new theories, on the other. In particular, we show...
Presentation
In this talk I will present some of the background to the project FraCaS which led to the FraCaS test suite. I will also talk about the MultiFraCaS project and some of our ideas for extending this in the future. The examples in the origi- nal test suite were created by semanticists. I will discuss a number of ways in which one could go about verify...
Article
Full-text available
This paper explores a connection between Romance and Greek on the one hand, and Bantu on the other. More specifically, we look at auxiliary placement in Rangi and clitic placement in Tobler Mussafia languages, with a special emphasis on Cypriot Greek, and argue that a common explanation for their distribution can be found once a move into a dynamic...
Chapter
Full-text available
Type theories, from the early days of Montague Semantics [27] to the recent work of using rich or modern type theories, have a long history of being employed as foundational languages of natural language semantics. In this introductory chapter, we will first describe and discuss the development of type theories as foundational languages of mathemat...
Article
Full-text available
Type theory was originally designed in order to overcome problems in the formalization of mathematics. It has made an immense contribution to the study of the foundations of mathematics, logic and computer science and has also played a central role in formal semantics for natural languages since the initial work of Richard Montague building on the...
Presentation
Full-text available
Presentation at INTERACTION2016
Presentation
Full-text available
Presentation at the 37th Annual Meeting of the Linguistics Department of Aristotle University
Article
Full-text available
This paper deals with the syntax of afterthoughts in Greek under a parsing-oriented perspective. The main claim is that afterthoughts can receive a straightforward explanation once we make the assumption that afterthoughts can be seen as answers to implicit questions. A formal syntactic account based on this assumption is put forth and its ability...
Article
Full-text available
In this paper, we show how a rich lexico-semantic networkwhich has been built using serious games, JeuxDeMots, can help us in grounding our semantic ontologies in doing formal semantics using rich or modern type theories (type theories within the tradition of Martin Löf). We discuss the issue of base types, adjectival and verbal types, hyperonymy/h...
Research
Full-text available
Slides for a talk given at the Centre for Linguistic Theory and Studies in Probability
Article
We argue that to reflect participant interactivity in conversational dialogue, the Christiansen & Chater (C&C) perspective needs a formal grammar framework capturing word-by-word incrementality, as in Dynamic Syntax, in which syntax is the incremental building of semantic representations reflecting real-time parsing dynamics. We demonstrate that, w...
Article
Full-text available
We argue that to reflect participant interactivity in conversational dialogue, the C&C perspective needs a formal grammar framework capturing word-by-word incrementality, as in Dynamic Syntax, in which syntax is the incremental building of semantic representations reflecting real-time parsing dynamics. We demonstrate that, with such formulation, sy...
Conference Paper
Full-text available
Dans cet article, nous présentons une utilisation des assistants des preuves pour traiter l'inférence en Language Naturel (NLI). D' abord, nous proposons d'utiliser les theories des types modernes comme langue dans laquelle traduire la sémantique du langage naturel. Ensuite, nous implémentons cette sémantique dans l'assistant de preuve Coq pour rai...
Chapter
Full-text available
In this article, the historical development of the clitic systems of Standard Modern, Cypriot and Pontic Greek are discussed. These three varieties not only present the whole range of variation one can find across clitic systems in Greek but furtermore derive from a common linguistic ancestor, i.e. Koine Greek. The goal of this paper is to show tha...
Conference Paper
Full-text available
This paper examines the distribution of two focus-associating operators in Greek, mono 'only' and additive ke 'too'. We establish the syntactic and prosodic factors that determine association with focus in this language. We argue that association with focus in Greek is subject to a syntactic requirement of sisterhood. We use this to probe into the...
Conference Paper
Full-text available
In this paper, polydefinites in Modern Greek are revisited from the perspective of Dynamic Syntax. The paper follows the idea set forth by Lekakou and Szendroi, 2012 according to which polydefinites are instances of apposition, in particular close apposition. Building on this idea, it is argued that loose' polydefinites, i.e. polydefnites that are...
Technical Report
Full-text available
In what follows, I discuss a number of cases of polydefinites that unexpectedly give rise to a non-restrictive interpretation in Modern Greek. It is shown that in most of the cases identified as problematic in the literature, a clear line of explanation can be pursued. 1. The case of eksipnos One of the well-known examples that have been used in th...
Conference Paper
Full-text available
In this paper, signatures are introduced to represent situations in formal semantics based on modern type theories. We study the notion of signature in more detail, presenting it formally and discussing its use in the representation of situations. In particular, the new forms of signature entries, subtyping entries and manifest entries, are formall...
Presentation
Full-text available
Slides for an advanced course taught at ESSLLI2014
Conference Paper
Full-text available
This paper is the first attempt to deal with aspects of the semantics of adverbs within a modern type-theoretical setting. A number of issues pertaining to the semantics of different classes of adverbs like verididality and intensionality will be discussed and further shown to be captured straightforwardly within a modern type-theoretical setting....
Conference Paper
Full-text available
In this paper, we study natural language inference based on the formal semantics in modern type theories (MTTs) and their implementations in proof-assistants such as Coq. To this end, the type theory UTT with coercive subtyping is used as the logical language in which natural lan-guage semantics is translated to, followed by the implementation of t...
Conference Paper
Full-text available
In this paper we discuss the semantics of adjectives from the perspective of a modern type theory (MTT) with an adequate subtyping mechanism. In an MTT, common nouns (CNs) can be interpreted as types and, in particular, CNs modified by intersective and subsective adjectives can be given semantics by means of Σ-types. However, an interpretation of C...