The purpose of this study is to develop an effective hybrid model for automatic document classification by combining statistical and semantic text vectorization techniques with machine learning algorithms. The methodology integrates Term Frequency–Inverse Document Frequency (TF-IDF) and Word2Vec embeddings with classifiers such as Support Vector Machine (SVM) and Random Forest. The proposed approach includes data preprocessing (tokenization, normalization, stop word removal, and lemmatization), feature extraction, model training, and evaluation using classification metrics such as accuracy, F1-score, Matthews Correlation Coefficient (MCC), and Cohen’s Kappa. Experimental results demonstrate that the Word2Vec + SVM model outperforms other configurations, achieving 90.2% accuracy and an F1-score of 82.52%, thus highlighting the advantage of incorporating semantic context into vector representation. The study concludes that hybrid methods combining TF-IDF and Word2Vec with robust classifiers improve both the precision and generalizability of document analysis models. Practical implications include potential applications in sentiment analysis, topic modeling, text classification for legal and healthcare domains, and multilingual contexts. This research provides a foundation for developing high-performance text analysis systems applicable to various real-world natural language processing tasks.

This paper proposes a hybrid, interpretable machine learning framework for the preliminary screening of soil macronutrients using Sentinel-2 and AgroLens data. This study aims not to replace laboratory analysis, but to test the feasibility of obtaining a useful proxy signal for estimating nitrogen (N), phosphorus (P), and potassium (K) content using a limited set of remote sensing and agricultural features. The developed pipeline includes data auditing, leakage control, feature engineering, train-only normalization, group-aware partitioning, baseline/SOTA model comparison, hybrid regression modeling, SHAP interpretation, and uncertainty assessment. The experiment used 4471 AgroLens observations and 126 features derived from Sentinel-2 spectral aggregates, vegetation indices, temporal characteristics, and crop-related parameters. The evaluation indicated that the proposed approach consistently improves forecasting quality relative to baseline models under reduced-input conditions. Linear relationships between target variables ranged from 0.14 to 0.17, while nonlinear relationships reached 0.23. SHAP analysis revealed significant contributions from vegetation indices, crop-specific interactions, and Sentinel-2 spectral channels. The findings support the applicability of the proposed framework for preliminary monitoring, prioritizing field surveys, and decision support in digital agriculture. Although an additional AgroLens control segment was used to assess the robustness of the study, the study did not include independent external validation of the data collected across different geographic or agro-climatic conditions.