Modelling of Learning Unnatural Language Quantifiers

Aktepe, Semih Can
The fact that all natural language quantifiers are conservative brings out the question of whether people have a predisposition to learn the conservative quantifiers over the non-conservative quantifiers. Many developmental and computational studies attempt to find an answer to this question, yet they fail to reach a consensus. This study tries to bring new insight into this debate by investigating the acquisition of four unnatural language quantifiers with a computational model and an eye-tracking experiment with a novel paradigm. The computational model makes some predictions about the effects of conservativity and complexity on quantifier learning. The predictions of the model are evaluated with a human subject eye- tracking experiment. This experiment employing the occluded referent paradigm provides us with what referents the participants utilize when discovering the meanings of the quantifiers. The results show that people have a bias for the most lexically-related referent rather than the conservativity bias, although that referent is not related when figuring out the meaning of the quantifiers. These findings open a new discussion about how people learn the meanings of the natural language quantifiers over the unnatural language quantifiers.


Automatic sense prediction of explicit discourse connectives in Turkish with the help of centering theory and morphosyntactic features
Çetin, Savaş; Zeyrek Bozşahin, Deniz; Department of Cognitive Sciences (2018)
Discourse connectives (and, but, however) are one of many means of keeping the discourse coherent. Discourse connectives are classified into groups based on their senses (expansion, contingency, etc.). They describe the semantic relationship of two discourse units. This study aims to build a machine learning system to predict the sense of explicit discourse connectives on the Turkish Discourse Bank data, which is manually gold-annotated. To do so, this study examines the effect of several features: i.e. tra...
Transformation of the State and Class Relations: Furthering Authoritarianism in Turkey
Topal Yılmaz, Aylin (null; 2017-11-09)
This paper aims to contribute to the debate on the transformation of state in different historically specific contexts by problematising the concept of authoritarian neoliberalism. It intends to do so by exploring the reorganisation of social forces and transformation of state power within an authoritarian state form as exemplified by the case of Turkey. It will also attempt to refresh class analysis in order to develop a better understanding of different modalities of reproduction of labour quite often wit...
Challenging the Liberal Establishment and Consolidating the Authoritarian Regime: Comparing Populism(s) in Contemporary Western Europe and Russia
Çelov, İgor; Deveci, Cem; Department of Political Science and Public Administration (2022-3-25)
The concept of populism has become particularly salient in the academic debates of recent years. Yet, there are few cross-regional studies of the populist phenomena. Comparisons of populism across qualitatively different polities are even fewer. The main reason behind this gap can be attributed to the fact that there is as much dispute about defining populism as there is about studying it, both of which contribute to the theoretical dissonance of populism studies. In this thesis, I attempt to bridge the gap...
Automatic sense prediction of implicit discourse relations in Turkish
Kurfalı, Murathan; Zeyrek Bozşahin, Deniz; Department of Cognitive Sciences (2016)
In discourse parsing, the sense prediction of the Implicit discourse relations poses the most significant challenge. The thesis aims to develop a supervised system to predict the sense of implicit discourse relations in Turkish Discourse Bank (TDB). In order to accomplish that goal, the discourse level annotations obtained from TDB are used. TDB follows the PDTB-2’s sense hierarchy and for all experiments within the current study, only CLASS senses are considered. As the primary experiment, the classifiers ...
Attention mechanisms for semantic few-shot learning
Baran, Orhun Buğra; Cinbiş, Ramazan Gökberk; İkizler-Cinbiş, Nazlı; Department of Computer Engineering (2021-9-1)
One of the fundamental difficulties in contemporary supervised learning approaches is the dependency on labelled examples. Most state-of-the-art deep architectures, in particular, tend to perform poorly in the absence of large-scale annotated training sets. In many practical problems, however, it is not feasible to construct sufficiently large training sets, especially in problems involving sensitive information or consisting of a large set of fine-grained classes. One of the main topics in machine learning...
Citation Formats
S. C. Aktepe, “Modelling of Learning Unnatural Language Quantifiers,” M.S. - Master Of Science Without Thesis, Middle East Technical University, 2022.