Papers by Senja Pollak

LLMSegm: Surface-level Morphological Segmentation Using Large Language Model (2024.lrec-main)

Copied to clipboard

Marko Pranjić, Marko Robnik-Šikonja, Senja Pollak

Challenge:	Existing approaches to morphological segmentation split word into its morphemes . LLMSegm is applicable in low-data settings and low-resourced languages .
Approach:	They propose a novel approach to surface-level morphological segmentation leveraging large language models.
Outcome:	The proposed method is applicable in low-data settings and low-resource languages.

Leveraging Contextual Embeddings for Detecting Diachronic Semantic Shift (2020.lrec-1)

Copied to clipboard

Matej Martinc, Petra Kralj Novak, Senja Pollak

Representation Learning

Challenge:	Existing methods for word embeddings have been used to model semantic relations with word embeds.
Approach:	They propose a method that leverages contextual embeddings for diachronic semantic shift detection by generating time specific word representations from BERT embedds.
Outcome:	The proposed method performs comparable to the current state-of-the-art without time consuming domain adaptation on large corpora.

Embeddings models for Buddhist Sanskrit (2022.lrec-1)

Copied to clipboard

Ligeia Lugli, Matej Martinc, Andraž Pelicon, Senja Pollak

Semantic Text Processing

Challenge:	Despite extensive scholarly endeavors, much uncertainty still surrounds this body of literature, especially regarding matters of chronology, authorship, compositional history.
Approach:	They propose a corpus of Buddhist texts, a general corpus and word similarity and word analogy datasets for embeddings models.
Outcome:	The proposed models perform better on semantic similarity and word analogy tasks than on contextual models.

A Computational Analysis of the Dehumanisation of Migrants from Syria and Ukraine in Slovene News Media (2024.lrec-main)

Copied to clipboard

Jaya Caporusso, Damar Hoogland, Mojca Brglez, Boshko Koloski, Matthew Purver, Senja Pollak

Challenge:	Dehumanisation involves the perception and/or treatment of a social group’s members as less than human.
Approach:	They propose to use a new sentiment resource to make it easier to transfer to other languages and to evaluate and use . they then apply the method to study attitudes to migration expressed in Slovene newspapers, and examine how this discourse changed between the 2015-16 migration crisis and the 2022-23 period following the war in Ukraine.
Outcome:	The proposed method is easier to transfer to other languages and evaluates . it combines zero-shot cross-lingual valence and arousal detection with statistical significance testing to examine attitudes to migration expressed in Slovene newspapers .

A Computational Framework to Identify Self-Aspects in Text (2025.acl-srw)

Copied to clipboard

Jaya Caporusso, Matthew Purver, Senja Pollak

Responsible & Trustworthy NLP

Challenge:	a Ph.D. proposal aims to identify Self-aspects in text, which are underexplored in natural language processing . many aspects of the Self align with psychological and other well-researched phenomena .
Approach:	They propose to develop a computational framework to identify Self-aspects in text . they will use an ontology of Self-facets and an annotated gold-standard dataset .
Outcome:	The proposed framework will evaluate discriminative models, generative large language models, embedding-based retrieval approaches against four main criteria: interpretability, ground-truth adherence, accuracy, and computational efficiency.

Reusable workflows for gender prediction (L18-1)

Copied to clipboard

Matej Martinc, Senja Pollak

Challenge:	Existing systems for author profiling (AP) modeling require extensive feature engineering and testing.
Approach:	They propose to implement a system for author profiling (AP) modeling that reduces the complexity and time of building a sophisticated model for a number of different AP tasks.
Outcome:	The proposed model achieves comparable results to state of the art models for cross-genre gender prediction, but lags when genre of test set is different from genre of train set.

CoSimLex: A Resource for Evaluating Graded Word Similarity in Context (2020.lrec-1)

Copied to clipboard

Carlos Santos Armendariz, Matthew Purver, Matej Ulčar, Senja Pollak, Nikola Ljubešić, Mark Granroth-Wilding

Semantic Text Processing

Challenge:	Existing methods to evaluate word embeddings ignore context and treat words in isolation.
Approach:	They propose to build a new word embeddings-based dataset that provides context-dependent similarity measures.
Outcome:	The proposed dataset provides context-dependent similarity measures and covers a well-resourced language (English) but a number of less-resource languages.

Denoising Labeled Data for Comment Moderation Using Active Learning (2024.lrec-main)

Copied to clipboard

Andraž Pelicon, Mladen Karan, Ravi Shekhar, Matthew Purver, Senja Pollak

Responsible & Trustworthy NLP

Challenge:	Large contextualized language models (LLMs) are becoming ubiquitous in natural language processing due to their performance and adaptability to diverse tasks.
Approach:	They propose to use active learning methods to denoise textual data for model training by sampling the most informative examples with noisy labels with active learning.
Outcome:	The proposed method reduces the cost of reannotation by reducing noise in noisy examples.

Thesis Proposal: Measuring Prejudice at Scale (2026.eacl-srw)

Copied to clipboard

Zoran Fijavž, Senja Pollak, Veronika Bajt

Text Classification

Challenge:	Existing methods for NLP fail to confirm construct validity, limiting the validity of the model.
Approach:	They propose to shift from categorical classification to comparative scaling of grounded constructs by using prompt optimization and distillation approaches.
Outcome:	The proposed pipeline is scalable for moving from categorical classification to theoretically grounded comparative measurement.

SEKE: Specialised Experts for Keyword Extraction (2025.findings-emnlp)

Copied to clipboard

Matej Martinc, Thi Hong Hanh Tran, Senja Pollak, Boshko Koloski

Information Extraction & Text Mining

Challenge:	Keyword extraction involves identifying the most descriptive words in a document . supervised keyword extraction is based on the mixture of experts (MoE) technique .
Approach:	They propose a supervised keyword extraction approach based on the mixture of experts technique . they use a learnable routing sub-network to direct information to specialised experts .
Outcome:	The proposed approach is based on the mixture of experts (MoE) technique . experts attend to each token and integrate it with a bidirectional long-term memory network .

Extracting and Analysing Metaphors in Migration Media Discourse: towards a Metaphor Annotation Scheme (2022.lrec-1)

Copied to clipboard

Ana Zwitter Vitez, Mojca Brglez, Marko Robnik Šikonja, Tadej Škvorc, Andreja Vezovnik, Senja Pollak

Challenge:	Using metaphors in media discourse is an increasingly researched topic . media are an important shaper of social reality and metaphors indicate how we think about issues through references to other things.
Approach:	They propose a neural transfer learning method for detecting metaphorical sentences in Slovene . scheme can be used for future metaphor annotations of other socially relevant topics .
Outcome:	The proposed method can be used for future metaphor annotations of other socially relevant topics.

Out of Thin Air: Is Zero-Shot Cross-Lingual Keyword Detection Better Than Unsupervised? (2022.lrec-1)

Copied to clipboard

Boshko Koloski, Senja Pollak, Blaž Škrlj, Matej Martinc

Low-Resource NLP

Challenge:	Keyword extraction is the task of retrieving words that are essential to the content of a document.
Approach:	They propose to use pretrained multilingual language models for zero-shot cross-lingual keyword extraction on low-resource languages with limited or no available labeled training data.
Outcome:	The proposed models outperform state-of-the-art unsupervised methods on low-resource languages with limited or no training data.