Advanced Multilingual Natural Language Processing in Industrial Digitalization: A Case Study of Autonomous Document Classification and ERP Integration in Manufacturing

AI-based order recognition module

In this work, presented is a technical description of a software module for automating customer order processing in an industrial company. Analyzed is the finding that manual processing of unstructured orders takes twenty to forty minutes per document and leads to errors when matching against a catalog of four thousand items. Investigated is the microservice architecture based on FastAPI, comprising a RAG service, telemetry collector, and an embedded widget for the Bitrix24 CRM system. Revealed is that hybrid search combining semantic vector representations and a lexical inverted index provides high accuracy in matching customer descriptions against catalog nomenclature. Studied is the application of a three-level text extraction strategy for documents in PDF, Excel, CSV, and ZIP formats with Tesseract OCR fallback. Determined is that text chunking with overlap and a multi-level LLM API invocation strategy enable reliable extraction of product items from orders of arbitrary volume. Established is that fine-tuning the E5-base model on a domain-specific corpus using triplet loss with hard negatives improves semantic search quality. Formed is an iterative accuracy improvement mechanism through a closed-loop telemetry collection and manager feedback system. Proposed is a two-stage nomenclature matching scheme with gate optimization and caching to accelerate processing of typical queries. Developed is integration with Bitrix24 and 1C systems automating the complete cycle from file upload to order creation in the accounting system. Substantiated is the advantage of the proposed approach over manual processing and classical full-text search in both speed and result quality. Presented is a description of validation results on real customer data confirming a three-to-four-fold reduction in order processing time.

LLM for the development of FCM

This article is about the development of a fuzzy cognitive map using a local large language model, and the model is thoroughly tested; Qwen2.5-32B is used and the data is extracted from hotel reviews from TripAdvisor and a fuzzy cognitive map is trained and evaluated.

of AI Based

Intelligent One-Gate System Based on Natural Language Processing for Enhancing Academic Information Services

The rapid digitalization of higher education requires academic information services that are fast, integrated, and accessible. Fragmented service channels, high administrative workloads, and slow response times remain persistent operational challenges. This study develops an Intelligent Transformer-Based One-Gate System that centralizes academic services by integrating text classification, abstractive summarization, and chatbot modules. Using a Research and Development approach, the system was built from 1,000 academic documents and service queries collected from FAQs, academic regulations, guides, and digital service archives. The corpus was cleaned, tokenized, encoded, and divided into training, validation, and testing subsets. BERT was applied for document and query classification, while BART and PEGASUS were evaluated for abstractive summarization using ROUGE metrics. The chatbot was assessed through a Likert-scale user acceptance survey. The results of this research show that the BERT classifier achieved 88% accuracy and an F1-score of 0.875, PEGASUS outperformed BART with ROUGE-1 = 0.74, ROUGE-2 = 0.66, and ROUGE-L = 0.71, and the chatbot achieved an average user satisfaction score of 82%. The main contribution of this research is an integrated transformer-based one-gate architecture that combines document routing, academic document summarization, and conversational assistance in a single service platform, offering practical value for reducing fragmented academic information services and methodological value as a reference model for higher education NLP implementation.

Requirement Classification and Prioritization Using Artificial Intelligence and Natural Language Processing

Requirement engineering is a foundational stage of the software development life cycle, and the accuracy with which requirements are classified directly influences downstream design, testing, and cost estimation. Software requirements are commonly separated into Functional Requirements (FR), which describe what a system must do, and Non-Functional Requirements (NFR), which describe how well the system must do it, covering attributes such as performance, security, and usability. When this separation is carried out manually, the process is slow, subjective, and prone to disagreement between analysts, particularly as project size grows. This paper presents a Requirement Classification and Prioritization Tool that combines Natural Language Processing (NLP) with machine learning to automate the FR/NFR decision. Requirement statements are cleaned and normalised, then represented numerically through two complementary techniques: Term FrequencyInverse Document Frequency (TF-IDF) and contextual BERT embeddings. Four classifiers-Logistic Regression, Support Vector Machine, Random Forest, and a BERT-based model-are trained and benchmarked on a dataset of 6,086 labelled requirement statements using accuracy, precision, recall, and F1-score. A Gradio-based interface allows a requirement to be submitted and its predicted category, confidence score, and a short explanation to be viewed immediately. The results indicate that transformer-based representations offer a modest but consistent improvement in contextual understanding over TF-IDF, while classical classifiers remain competitive and considerably cheaper to train.

AI-Based Assistant for Generating and Analyzing Legal Documents

The legal industry has been heavily impacted by fast-paced digitization, but the complete volume of complex legal documents, technical jargon, and arcane procedures continues to limit access to legal knowledge for both professionals and the general public. This highlights the need for more powerful, AI-based solutions for the retrieval and comprehension of legal information. The state-of-the-art AI and natural language processing (NLP)-based legal assistance systems are designed to increase accessibility, but are limited in their practical impact, scalability, and explainability. In this paper, we present an AI-assisted tool for the generation and analysis of legal documents. The system is based on a fine-tuned Distil-BERT model for effective classification of legal documents, semantic comprehension, and user-centric legal information support. The framework fulfils data preprocessing, transformer-based feature extraction, and supervised classification in an end-to-end manner, and can be embedded in legal service applications. The experimental results on a legal document dataset show that the proposed Distil-BERT model achieved an accuracy of 82.01% and a mean F1 of 84.32%. These results demonstrate that distillation-based knowledge transformers can achieve better generalization with less complexity, which is a desirable property for scalable solutions in the legal AI context. Further, we present important ethical and deployment factors to consider (e.g., data privacy, legal liability of AI outputs, and algorithmic bias mitigation). In sum, the proposed system takes a step toward intuition-friendly, efficient, and ethically conscious AI-supported legal aid that furthers the digital transformation of legal services.