Literary translation is a difficult task that not only requires semantic accuracy but also stylistic richness and lexical diversity. Pretrained and supervised fine-tuned Large Language Models (LLMs) can over-rely on safe vocabulary choices, leading to translations that lack lexical variety. To address this problem, we propose a novel diversity-aware multi-objective Group Relative Policy Optimization (GRPO) framework that pushes the limits of open-source translation quality while increasing lexical diversity. We introduce two diversity-aware reward mechanisms, a Leave-One-Out (LOO) marginal contribution reward and a Self-BLEU penalty, balanced alongside neural quality metrics (COMET), lexical overlap (BLEU), and structural constraints. Through experiments on Turkish-English and German-English using Qwen3-14B, we show that our diversity-aware reinforcement learning approach successfully enhances lexical richness alongside translation quality. Our models achieve state-of-the-art open-source performance in literary translation, bridging the gap with leading commercial systems and demonstrating that policy optimization can effectively steer LLMs toward high-quality, lexically diverse outputs.

Pretrained language models and large language models are increasingly used to assist in a great variety of natural language tasks. In this work, we explore their use in evaluating the quality of alternative corpus annotation schemes. For this purpose, we analyze two alternative annotations of the Turkish BOUN treebank, versions 2.8 and 2.11, in the Universal Dependencies framework using large language models. Using a suitable prompt generated using treebank annotations, large language models are used to recover the surface forms of sentences. Based on the idea that the large language models capture the characteristics of the languages, we expect that the better annotation scheme would yield the sentences with higher success. The experiments conducted on a subset of the treebank show that the new annotation scheme (2.11) results in a successful recovery percentage of about 2 points higher. All the code developed for this work is available at https://github.com/boun-tabi/eval-ud .

Although machine translation systems are mostly designed to serve in the general domain, there is a growing tendency to adapt these systems to other domains like literary translation. In this paper, we focus on English-Turkish literary translation and develop machine translation models that take into account the stylistic features of translators. We fine-tune a pre-trained machine translation model by the manually-aligned works of a particular translator. We make a detailed analysis of the effects of manual and automatic alignments, data augmentation methods, and corpus size on the translations. We propose an approach based on stylistic features to evaluate the style of a translator in the output translations. We show that the human translator style can be highly recreated in the target machine translations by adapting the models to the style of the translator.

Abstract The tremendous amount of increase in the number of documents available on the Web has turned finding the relevant piece of information into a challenging, tedious, and time-consuming activity. Accordingly, automatic text summarization has become an important field of study by gaining significant attention from the researchers. Lately, with the advances in deep learning, neural abstractive text summarization with sequence-to-sequence (Seq2Seq) models has gained popularity. There have been many improvements in these models such as the use of pretrained language models (e.g., GPT, BERT, and XLM) and pretrained Seq2Seq models (e.g., BART and T5). These improvements have addressed certain shortcomings in neural summarization and have improved upon challenges such as saliency, fluency, and semantics which enable generating higher quality summaries. Unfortunately, these research attempts were mostly limited to the English language. Monolingual BERT models and multilingual pretrained Seq2Seq models have been released recently providing the opportunity to utilize such state-of-the-art models in low-resource languages such as Turkish. In this study, we make use of pretrained Seq2Seq models and obtain state-of-the-art results on the two large-scale Turkish datasets, TR-News and MLSum, for the text summarization task. Then, we utilize the title information in the datasets and establish hard baselines for the title generation task on both datasets. We show that the input to the models has a substantial amount of importance for the success of such tasks. Additionally, we provide extensive analysis of the models including cross-dataset evaluations, various text generation options, and the effect of preprocessing in ROUGE evaluations for Turkish. It is shown that the monolingual BERT models outperform the multilingual BERT models on all tasks across all the datasets. Lastly, qualitative evaluations of the generated summaries and titles of the models are provided.

Success of neural networks in natural language processing has paved the way for neural machine translation (NMT), which rapidly became the mainstream approach in machine translation. Significant improvement in translation performance has been achieved with breakthroughs such as encoder-decoder networks, attention mechanism, and Transformer architecture. However, the necessity of large amounts of parallel data for training an NMT system and rare words in translation corpora are issues yet to be overcome. In this article, we approach NMT of the low-resource Turkish-English language pair. We employ state-of-the-art NMT architectures and data augmentation methods that exploit monolingual corpora. We point out the importance of input representation for the morphologically rich Turkish language and make a comprehensive analysis of linguistically and non-linguistically motivated input segmentation approaches. We prove the effectiveness of morphologically motivated input segmentation for the Turkish language. Moreover, we show the superiority of the Transformer architecture over attentional encoder-decoder models for the Turkish-English language pair. Among the employed data augmentation approaches, we observe back-translation to be the most effective and confirm the benefit of increasing the amount of parallel data on translation quality. This research demonstrates a comprehensive analysis on NMT architectures with different hyperparameters, data augmentation methods, and input representation techniques, and proposes ways of tackling the low-resource setting of Turkish-English NMT.

Relation Extraction (RE) is the task of finding the relation between entities in a plain text. As the length of the sentences increases, finding the relation becomes more challenging. The shortest dependency path (SDP) between two entities, obtained by traversing the terms in the dependency tree of a sentence, provides a view focused on the entities by pruning noisy words. In the supervised form of the relation extraction task, Relation Classification, the state-of-the-art methods generally integrate a pre-trained language model (PLM) into their approaches. However, none of them incorporates the shortest dependency paths to the best of our knowledge.This paper investigates the effects of using shortest dependency paths with pre-trained language models by taking the R-BERT relation classification model as the baseline and building upon it. Our novel approach enhances the baseline model by adding the sequence representation of the shortest dependency path between entities, collected from PLMs, as an additional embedding. In the experiments, we evaluated the proposed model’s performance for each combination of SDPs generated from Stanford, HPSG, and LAL dependency parsers with BERT and XLNet PLMs in two datasets, SemEval-2010 Task 8 and TACRED. We improved the baseline model by absolute 1.41% and 3.60% scores, increasing the rankings of the model from 8thto 7thand from 18thto 7thin SemEval-2010 Task 8 and TACRED, respectively.

Abstract Although many studies on sentiment analysis have been carried out for widely spoken languages, this topic is still immature for Turkish. Most of the works in this language focus on supervised models, which necessitate comprehensive annotated corpora. There are a few unsupervised methods, and they utilize sentiment lexicons either built by translating from English lexicons or created based on corpora. This results in improper word polarities as the language and domain characteristics are ignored. In this paper, we develop unsupervised (domain-independent) and semi-supervised (domain-specific) methods for Turkish, which are based on a set of antonym word pairs as seeds. We make a comprehensive analysis of supervised methods under several feature weighting schemes. We then form ensemble of supervised classifiers and also combine the unsupervised and supervised methods. Since Turkish is an agglutinative language, we perform morphological analysis and use different word forms. The methods developed were tested on two datasets having different styles in Turkish and also on datasets in English to show the portability of the approaches across languages. We observed that the combination of the unsupervised and supervised approaches outperforms the other methods, and we obtained a significant improvement over the state-of-the-art results for both Turkish and English.

Large annotated datasets in NLP are overwhelmingly in English.This is an obstacle to progress in other languages.Unfortunately, obtaining new annotated resources for each task in each language would be prohibitively expensive.At the same time, commercial machine translation systems are now robust.Can we leverage these systems to translate Englishlanguage datasets automatically?In this paper, we offer a positive response for natural language inference (NLI) in Turkish.We translated two large English NLI datasets into Turkish and had a team of experts validate their translation quality and fidelity to the original labels.Using these datasets, we address core issues of representation for Turkish NLI.We find that in-language embeddings are essential and that morphological parsing can be avoided where the training set is large.Finally, we show that models trained on our machinetranslated datasets are successful on humantranslated evaluation sets.We share all code, models, and data publicly.

Morphological information is important for many sequence labeling tasks in Natural Language Processing (NLP). Yet, existing approaches rely heavily on manual annotations or external software to capture this information. In this study, we propose using subword contextual embeddings for languages with rich morphology. Evaluated on Dependency Parsing (DEP) and Named Entity Recognition (NER) tasks, which are shown to benefit highly from morphological information, subword contextual embeddings consistently outperformed other approaches on all languages tested (Hungarian, Finnish, Czech and Turkish). Our proposed method enables achieving state-of-the-art results with little annotation requirements compared to the previous work. Besides, the novel network architecture we propose, coupled with a Bayesian hyperparameter optimization suite, achieved state-of-the-art results for both tasks for the Turkish language. Finally, we experimented with different multi-task learning architectures to analyze the effect of jointly learning the two tasks.

As the access to information in the education domain increases, new technologies are developing for school children. However, deaf and dumb children still have limited access to the information, especially in their school lives. One of the most important reasons for this problem is the lack of studies in the Sign Language domain. In this paper, we propose a novel method for translation from Turkish to Turkish Sign Language for primary school students using the statistical machine translation approach. To the best of our approach, this is the first work that applies statistical translation to Turkish Sign Language. A parallel corpus is compiled from the books published by Ministry of National Education of Turkey. The results of the system were tested using different evaluation metrics. We observe that the results obtained are motivating for new studies.

Sign language is the primary tool of communication for deaf and mute people. It employs hand gestures, facial expressions, and body movements to state a word or a phrase. Like spoken languages, sign languages also vary among the regions and the cultures. The aim of this study is to implement a machine translation system to convert Turkish spoken language into Turkish Sign Language (TID). The advantages of rule-based and statistical machine translation techniques are combined into a hybrid translation system.

Abstract This work proposes a sequential tagger for named entity recognition in morphologically rich languages. Several schemes for representing the morphological analysis of a word in the context of named entity recognition are examined. Word representations are formed by concatenating word and character embeddings with the morphological embeddings based on these schemes. The impact of these representations is measured by training and evaluating a sequential tagger composed of a conditional random field layer on top of a bidirectional long short-term memory layer. Experiments with Turkish, Czech, Hungarian, Finnish and Spanish produce the state-of-the-art results for all these languages, indicating that the representation of morphological information improves performance.

Previous studies have shown that linguistic features of a word such as possession, genitive or other grammatical cases can be employed in word representations of a named entity recognition (NER) tagger to improve the performance for morphologically rich languages. However, these taggers require external morphological disambiguation (MD) tools to function which are hard to obtain or non-existent for many languages. In this work, we propose a model which alleviates the need for such disambiguators by jointly learning NER and MD taggers in languages for which one can provide a list of candidate morphological analyses. We show that this can be done independent of the morphological annotation schemes, which differ among languages. Our experiments employing three different model architectures that join these two tasks show that joint learning improves NER performance. Furthermore, the morphological disambiguator’s performance is shown to be competitive.

Abstract Motivation Despite recent advances in algorithms design to characterize structural variation using high-throughput short read sequencing (HTS) data, characterization of novel sequence insertions longer than the average read length remains a challenging task. This is mainly due to both computational difficulties and the complexities imposed by genomic repeats in generating reliable assemblies to accurately detect both the sequence content and the exact location of such insertions. Additionally, de novo genome assembly algorithms typically require a very high depth of coverage, which may be a limiting factor for most genome studies. Therefore, characterization of novel sequence insertions is not a routine part of most sequencing projects. There are only a handful of algorithms that are specifically developed for novel sequence insertion discovery that can bypass the need for the whole genome de novo assembly. Still, most such algorithms rely on high depth of coverage, and to our knowledge there is only one method (PopIns) that can use multi-sample data to “collectively” obtain a very high coverage dataset to accurately find insertions common in a given population. Result Here, we present Pamir, a new algorithm to efficiently and accurately discover and genotype novel sequence insertions using either single or multiple genome sequencing datasets. Pamir is able to detect breakpoint locations of the insertions and calculate their zygosity (i.e. heterozygous versus homozygous) by analyzing multiple sequence signatures, matching one-end-anchored sequences to small-scale de novo assemblies of unmapped reads, and conducting strand-aware local assembly. We test the efficacy of Pamir on both simulated and real data, and demonstrate its potential use in accurate and routine identification of novel sequence insertions in genome projects. Availability and implementation Pamir is available at https://github.com/vpc-ccg/pamir. Supplementary information Supplementary data are available at Bioinformatics online.

Abstract In this paper, we study the problem of structural analysis of Web documents aiming at extracting the sectional hierarchy of a document. In general, a document can be represented as a hierarchy of sections and subsections with corresponding headings and subheadings. We developed two machine learning models: heading extraction model and hierarchy extraction model. Heading extraction was formulated as a classification problem whereas a tree-based learning approach was employed in hierarchy extraction. For this purpose, we developed an incremental learning algorithm based on support vector machines and perceptrons. The models were evaluated in detail with respect to the performance of the heading and hierarchy extraction tasks. For comparison, a baseline rule-based approach was used that relies on heuristics and HTML document object model tree processing. The machine learning approach, which is a fully automatic approach, outperformed the rule-based approach. We also analyzed the effect of document structuring on automatic summarization in the context of Web search. The results of the task-based evaluation on TREC queries showed that structured summaries are superior to unstructured summaries both in terms of accuracy and user ratings, and enable the users to determine the relevancy of search results more accurately than search engine snippets.

Collecting data through a game with a purpose (GWAP) has become a popular approach due to its numerous benefits. However, lack of diversity in game mechanisms puts some limitations on the areas of application of GWAPs. In this paper, we introduce a new two-phase mechanism for collecting data via human — based computation games. In the first phase, the players are provided with an object and asked to use that object within the domain of the game. In the second phase the players are provided with randomly selected data produced by the other players in the first phase and asked to evaluate them. A new game called “Dil Cambazı” that collects sentences containing passivized intransitive Turkish verbs using this mechanism is introduced and cases where input-evaluation mechanism is most useful are discussed.

This paper introduces two complementary language modeling approaches for morphologically rich languages aiming to alleviate out-of-vocabulary (OOV) word problem and to exploit morphology as a knowledge source. The first model, morpholexical language model, is a generative$n$-gram model, where modeling units are lexical-grammatical morphemes instead of commonly used words or statistical sub-words. This paper also proposes a novel approach for integrating the morphology into an automatic speech recognition (ASR) system in the finite-state transducer framework as a knowledge source. We accomplish that by building a morpholexical search network obtained by the composition of lexical transducer of a computational lexicon with a morpholexical language model. The second model is a linear reranking model trained discriminatively with a variant of the perceptron algorithm using morpholexical features. This variant of the perceptron algorithm, WER-sensitive perceptron, is shown to perform better for reranking$n$-best candidates obtained with the generative model. We apply the proposed models in Turkish broadcast news transcription task and give experimental results. The morpholexical model leads to an elegant morphology-integrated search network with unlimited vocabulary. Thus, it is highly effective in alleviating OOV problem and improves the word error rate (WER) over word and statistical sub-word models by 1.8% and 0.4% absolute, respectively. The discriminatively trained morpholexical model further improves the WER of the system by 0.8% absolute.

This paper explores rich morphological and novel n-best-list features for reranking automatic speech recognition hypotheses. The morpholexical features are defined over the morphological features obtained by using an n-gram language model over lexical and grammatical morphemes in the first-pass. The n-best-list features for each hypothesis are defined using that hypothesis and other alternate hypotheses in an n-best list. Our methodology is to align each hypothesis with other hypotheses one by one using minimum edit distance alignment. This gives us a set of edit operations - substitution, addition and deletion as seen in these alignments. These edit operations constitute our n-best-list features as indicator features. The reranking model is trained using a word error rate sensitive averaged perceptron algorithm introduced in this paper. The proposed methods are evaluated on a Turkish broadcast news transcription task. The baseline systems are word and statistical sub-word systems which also employ morphological features for reranking. We show that morpholexical and n-best-list features are effective in improving the accuracy of the system (0.8%).

We explore morphology-based and sub-word language modeling approaches proposed for morphologically rich languages, and evaluate and contrast them for Turkish broadcast news transcription task. In addition, as a morphology-based model, we improve our previously proposed morphology-integrated model for automatic speech recognition. This model is built by composing the finite-state transducer of the morphological parser with a language model over lexical morphemes. This approach provides a morphology-integrated search network with an unlimited vocabulary, generating only valid word forms while reducing the out-of-vocabulary rate and hence improving the word error rate. We also analyze the effect of morpho-tactics and morphological disambiguation on the speech recognition accuracy for the morphology-integrated model. The improved morphology-integrated model performs better than statistically derived sub-word models with added benefit of generating morpho-syntactic and semantic features.

This paper proposes a novel approach to integrate the morphology as a model into an automatic speech recognition (ASR) system for morphologically rich languages. The high out-of-vocabulary (OOV) word rates have been a major challenge for ASR in morphologically productive languages. The standard approach to this problem has been to shift from words to sub-word units in language modeling, and the only change to the system is in the language model estimated over these units. In contrast, we propose to integrate the morphology as other any knowledge source - such as the lexicon, and the language model- directly into the search network. The morphological parser for a language, implemented as a finite-state lexical transducer, can be considered as a computational lexicon. The computational lexicon represents a dynamic vocabulary in contrast to a static vocabulary generally used for ASR. We compose the transducer for this computational lexicon with a statistical language model over lexical morphemes to obtain a morphology-integrated search network. The resulting search network generates only grammatical word forms and improves the recognition accuracy due to reduced OOV rate. We give experimental results for Turkish broadcast news transcription, and show that it outperforms the 50K and 100K vocabulary word models while the 200K vocabulary word model is slightly better.

Sign analysis is a technique for deriving the behavior of a system when it is subjected to a disturbance. the technique is for systems whose mathematical models are of the form of algebraic equations. the technique is based on generating sign combinations of total differentials of system parameters in table form from closed form algebraic functions which model the system. Sign combinations are then analyzed to predict system behavior with respect to a disturbance from the equilibrium state. the technique may be applied to total differentials of gains as well. © 1992 John Wiley & Sons, Inc.