This study introduces a novel Natural Language Processing pipeline that enhances customer service efficiency at Telekom Srbija, a leading Serbian telecommunications company, through automated email topic detection and labelling. Central to the pipeline is BERTopic, a modular architecture that allows unsupervised topic modelling. After a series of preprocessing and post-processing steps, we assign one of 12 topics and several additional labels to incoming emails, allowing customer service to filter and access them through a custom-made application. The model's performance was evaluated by assessing the speed and correctness of the automatically assigned topics across a test dataset of 100 customer emails. The pipeline shows broad applicability across languages, particularly for those that are low-resourced and morphologically rich. The system now operates in the company's production environment, streamlining customer service operations through automated email classification.

Recent advancements in the field of natural language processing (NLP) and especially large language models (LLMs) and their numerous applications have brought research attention to design of different document processing tools and enhancements in the process of document archiving, search and retrieval. Domain of official, legal documents is especially interesting due to vast amount of data generated on the daily basis, as well as the significant community of interested practitioners (lawyers, law offices, administrative workers, state institutions and citizens). Providing efficient ways for automation of everyday work involving legal documents is therefore expected to have significant impact in different fields. In this work we present one LLM based solution for Named Entity Recognition (NER) in the case of legal documents written in Serbian language. It leverages on the pre-trained bidirectional encoder representations from transformers (BERT), which had been carefully adapted to the specific task of identifying and classifying specific data points from textual content. Besides novel dataset development for Serbian language (involving public court rulings), presented system design and applied methodology, the paper also discusses achieved performance metrics and their implications for objective assessment of the proposed solution. Performed cross-validation tests on the created manually labeled dataset with mean $F_1$ score of 0.96 and additional results on the examples of intentionally modified text inputs confirm applicability of the proposed system design and robustness of the developed NER solution.

Split Learning (SL) recently emerged as an efficient paradigm for distributed Machine Learning (ML) suitable for the Internet Of Things (IoT)-Cloud systems. However, deploying SL on resource-constrained edge IoT platforms poses a significant challenge in terms of balancing the model performance against the processing, memory, and energy resources. In this work, we present a practical study of deploying SL framework on a real-world Field-Programmable Gate Array (FPGA)-based edge IoT platform. We address the SL framework applied to a time-series processing model based on Recurrent Neural Networks (RNNs). Set in the context of river water quality monitoring and using real-world data, we train, optimize, and deploy a Long Short-Term Memory (LSTM) model on a given edge IoT FPGA platform in different SL configurations. Our results demonstrate the importance of aligning design choices with specific application requirements, whether it is maximizing speed, minimizing power, or optimizing for resource constraints.

Given the exponential advancement in AI technologies and the potential escalation of harmful effects from recommendation systems, it is crucial to simulate and evaluate these effects early on. Doing so can help prevent possible damage to both societies and technology companies. This paper introduces the Recommender Systems LLMs Playground (RecSysLLMsP), a novel simulation framework leveraging Large Language Models (LLMs) to explore the impacts of different content recommendation setups on user engagement and polarization in social networks. By creating diverse AI agents (AgentPrompts) with descriptive, static, and dynamic attributes, we assess their autonomous behaviour across three scenarios: Plurality, Balanced, and Similarity. Our findings reveal that the Similarity Scenario, which aligns content with user preferences, maximizes engagement while potentially fostering echo chambers. Conversely, the Plurality Scenario promotes diverse interactions but produces mixed engagement results. Our study emphasizes the need for a careful balance in recommender system designs to enhance user satisfaction while mitigating societal polarization. It underscores the unique value and challenges of incorporating LLMs into simulation environments. The benefits of RecSysLLMsP lie in its potential to calculate polarization effects, which is crucial for assessing societal impacts and determining user engagement levels with diverse recommender system setups. This advantage is essential for developing and maintaining a successful business model for social media companies. However, the study's limitations revolve around accurately emulating reality. Future efforts should validate the similarity in behaviour between real humans and AgentPrompts and establish metrics for measuring polarization scores.

With the expansion of neural networks, such as large language models, humanity is exponentially heading towards superintelligence. As various AI systems are increasingly integrated into the fabric of societies-through recommending values, devising creative solutions, and making decisions-it becomes critical to assess how these AI systems impact humans in the long run. This research aims to contribute towards establishing a benchmark for evaluating the sentiment of various Large Language Models in socially importan issues. The methodology adopted was a Likert scale survey. Seven LLMs, including GPT-4 and Bard, were analyzed and compared against sentiment data from three independent human sample populations. Temporal variations in sentiment were also evaluated over three consecutive days. The results highlighted a diversity in sentiment scores among LLMs, ranging from 3.32 to 4.12 out of 5. GPT-4 recorded the most positive sentiment score towards AGI, whereas Bard was leaning towards the neutral sentiment. The human samples, contrastingly, showed a lower average sentiment of 2.97. The temporal comparison revealed differences in sentiment evolution between LLMs in three days, ranging from 1.03% to 8.21%. The study's analysis outlines the prospect of potential conflicts of interest and bias possibilities in LLMs' sentiment formation. Results indicate that LLMs, akin to human cognitive processes, could potentially develop unique sentiments and subtly influence societies' perceptions towards various opinions formed within the LLMs.

The significance of distributed learning and inference algorithms in Internet of Things (IoT) network is growing since they flexibly distribute computation load between IoT devices and the infrastructure, enhance data privacy, and minimize latency. However, a notable challenge stems from the influence of communication channel conditions on their performance. In this work, we introduce COMSPLIT: a novel communication-aware design for split learning (SL) and inference paradigm tailored to processing time series data in IoT networks. COMSPLIT provides a versatile framework for deploying adaptable SL in IoT networks affected by diverse channel conditions. In conjunction with the integration of an early-exit strategy, and addressing IoT scenarios containing devices with heterogeneous computational capabilities, COMSPLIT represents a comprehensive design solution for communication-aware SL in IoT networks. Numerical results show superior performance of COMSPLIT compared to vanilla SL approaches (that assume ideal communication channel), demonstrating its ability to offer both design simplicity and adaptability to different channel conditions.

The robustness of AI-content detection models against sophisticated adversarial strategies, such as paraphrasing or word switching, is a rising concern in natural language generation (NLG) applications. This study proposes ToBlend, a novel token-level ensemble text generation method to challenge the robustness of current AI-content detection approaches by utilizing multiple sets of candidate generative large language models (LLMs). By randomly sampling token(s) from candidate LLMs sets, we find ToBlend significantly drops the performance of most mainstream AI-content detection methods. We evaluate the text quality produced under different ToBlend settings based on annotations from experienced human experts. We proposed a fine-tuned Llama3.1 model to distinguish the ToBlend generated text more accurately. Our findings underscore our proposed text generation approach's great potential in deceiving and improving detection models. Our datasets, codes, and annotations are open-sourced.

Establishing and maintaining 5G mmWave vehicular connectivity poses a significant challenge due to high user mobility that necessitates frequent triggering of beam switching procedures. Departing from reactive beam switching based on the user device channel state feedback, proactive beam switching prepares in advance for upcoming beam switching decisions by exploiting accurate channel state information (CSI) prediction. In this paper, we develop a framework for autonomous self-trained CSI prediction for mmWave vehicular users where a base station (gNB) collects and labels a dataset that it uses for training recurrent neural network (RNN)-based CSI prediction model. The proposed framework exploits the CSI feedback from vehicular users combined with overhearing the C-V2X cooperative awareness messages (CAMs) they broadcast. We implement and evaluate the proposed framework using deepMIMO dataset generation environment and demonstrate its capability to provide accurate CSI prediction for 5G mmWave vehicular users. CSI prediction model is trained and its capability to provide accurate CSI predictions from various input features are investigated.

The Closest String Problem is an NP-hard problem that aims to find a string that has the minimum distance from all sequences that belong to the given set of strings. Its applications can be found in coding theory, computational biology, and designing degenerated primers, among others. There are efficient exact algorithms that have reached high-quality solutions for binary sequences. However, there is still room for improvement concerning the quality of solutions over DNA and protein sequences. In this paper, we introduce a three-stage algorithm that comprises the following process: first, we apply a novel alphabet pruning method to reduce the search space for effectively finding promising search regions. Second, a variant of beam search to find a heuristic solution is employed. This method utilizes a newly developed guiding function based on an expected distance heuristic score of partial solutions. Last, we introduce a local search to improve the quality of the solution obtained from the beam search. Furthermore, due to the lack of real-world benchmarks, two real-world datasets are introduced to verify the robustness of the method. The extensive experimental results show that the proposed method outperforms the previous approaches from the literature.

In this paper, we introduce the VerifAI project, a pioneering open-source scientific question-answering system, designed to provide answers that are not only referenced but also automatically vetted and verifiable. The components of the system are (1) an Information Retrieval system combining semantic and lexical search techniques over scientific papers (PubMed), (2) a Retrieval-Augmented Generation (RAG) module using fine-tuned generative model (Mistral 7B) and retrieved articles to generate claims with references to the articles from which it was derived, and (3) a Verification engine, based on a fine-tuned DeBERTa and XLM-RoBERTa models on Natural Language Inference task using SciFACT dataset. The verification engine cross-checks the generated claim and the article from which the claim was derived, verifying whether there may have been any hallucinations in generating the claim. By leveraging the Information Retrieval and RAG modules, Verif.ai excels in generating factual information from a vast array of scientific sources. At the same time, the Verification engine rigorously double-checks this output, ensuring its accuracy and reliability. This dual-stage process plays a crucial role in acquiring and confirming factual information, significantly enhancing the information landscape. Our methodology could significantly enhance scientists' productivity, concurrently fostering trust in applying generative language models within scientific domains, where hallucinations and misinformation are unacceptable.