]]> The development of the Student Information System followed a step-by-step, practical approach. The project began with the design and implementation of the frontend using HTML, CSS, and JavaScript. The goal was to first create a functional and user-friendly interface for core pages such as login, dashboard, student list, course enrollment, and grade overview. Special attention was given to layout, usability, and responsiveness to ensure that the interface would meet the needs of different user roles.
Once the frontend was in place, backend logic was implemented using PHP. This phase involved connecting the visual components to a MySQL database and enabling dynamic data handling. Each feature created in the frontend — such as forms for registration, login, and grade entry — was connected to PHP scripts that processed the data, performed validations, and communicated with the database.
A relational database structure was created to support the application’s core entities: users, students, courses, and grades. SQL queries were used within PHP to retrieve, insert, update, and delete records, depending on the user’s actions. Session-based authentication was used to control access and ensure that each user could only interact with the system according to their assigned role (admin, professor, or student).
The system was developed iteratively, starting with smaller components and gradually building up to more complex functionality. Each step was tested using sample data to verify that both the frontend and backend behaved as expected. This phased approach helped identify and fix issues early, while also making it easier to expand the system with additional features.
The results of the project indicate that the proposed system significantly improves the accessibility and organization of student data, reduces the potential for errors, and simplifies administrative tasks. Faculty members can easily enter and update grades, while students have real-time access to their academic progress. The system provides a more transparent and efficient way of managing academic information, ultimately enhancing the communication between all parties involved. In conclusion, this Student Information System demonstrates how digital solutions can modernize and optimize traditional academic processes in an effective and scalable manner.
]]> The front end is built with React and Leaflet (Leaflet.draw over OpenStreetMap tiles), allowing workers to define one or more work areas by drawing circles, rectangles, and polygons. On the server, a NestJS, TypeORM, PostgreSQL/PostGIS stack normalizes incoming GeoJSON to SRID 4326 (WGS84) and consolidates all shapes into a canonical MultiPolygon with GiST indexing for efficient spatial queries [1].
Security is managed through a production-grade authentication subsystem, including OTP email verification, Google OAuth, and short-lived JWT access tokens (RS256/ES256) carrying a kid header mapped to a database Keystore and Signing Key tables. Refresh tokens are device-scoped, hashed, and rotated on use, while a public JWKS route exposes verification keys for third-party consumers. Role-scoped access is enforced via modular guards, with session-based authentication planned for first-party web sessions.
The matching system blends relevance and fairness by combining distance containment, price fit, category/competency match, listing freshness, and Bayesian worker rating into a configurable score. The query builder pre-filters by geometry and competencies before ranking, targeting sub-500 ms feed latency under spatial

indexing. A focused test suite with Jest unit tests validates token/key rotation, keystore lifecycle, query-builder filters, and geospatial utilities.
Operational deployment is packaged for Dokploy (API, database, secrets) with optional Tailscale access, while OpenAPI/Swagger documentation ensures discoverability.
Radesso contributes a cohesive blueprint—and a working core—for a localized job-seeking platform where workers proactively seek orders. The near-term roadmap includes reviews and appeals, enhanced admin moderation, and telemetry, positioning Radesso to evolve into a trusted, geospatially precise hub for local work.
]]> The platform's user-friendly design was created using React, powered by Vite.js, while the process going on behind the scenes were developed using Spring-boot and protected using Spring-security. The two sides communicate through RESTful API and the data is stored in a MySQL database. Key functionalities include searching for, filtering and booking tickets. When the trip is completed users can leave a review, and of course they cancel their reservation if they wish.
Ultimately, this project showcases how digital tools can simplify everyday tasks, offering an accessible solution for trip reservation that has the potential to improve the travel experience for all age groups.
]]> ]]>
The problem addressed in this project is the lack of centralized, real-time access to personal medical information for patients, and inefficient manual processes in health institutions. To solve this, the application was developed using a modular architecture consisting of a frontend interface built with HTML, CSS, and JavaScript, and a backend implemented in PHP following the RESTful API principles. The backend is organized into DAO, service, and route layers, and communicates with a MySQL database.

Advanced features include two-factor authentication (OTP via email or SMS), Google OAuth login, hCaptcha bot protection, and secure password checks using the HaveIBeenPwned API. Role-based access control (RBAC) was implemented to ensure that patients and doctors access only the data relevant to their role. The application also includes validation of phone numbers and email domains (TLD and MX records), and follows modern web security practices.
The result is a responsive and secure web application that streamlines interaction between patients and healthcare providers, improves the organization of medical data, and enhances user experience. Testing has confirmed that the system performs reliably across different scenarios, and offers a foundation for potential integration with official e-health systems
]]> The research focuses on evaluating various classical machine learning algorithms for sentiment classification in financial texts. A publicly available dataset combining FIQA and Financial PhraseBank sources was used, containing over 5,800 labeled financial sentences. Data preprocessing steps included cleaning, tokenization, stopword removal, and lemmatization. Exploratory data analysis was conducted to understand sentiment distribution, text length patterns, and word frequencies.
Sentiment labels were encoded numerically to serve as target variables in classification models. A range of traditional machine learning algorithms was implemented and assessed to explore their suitability for sentiment analysis in the financial domain. Evaluation metrics including accuracy, precision, recall, and F1-score were used to assess model performance.
Preliminary results indicate that traditional machine learning models can effectively classify sentiment in financial texts, especially when supported by proper text preprocessing. Furthermore, integrating sentiment analysis with financial data shows potential for improving market forecasting accuracy.


This work contributes to the growing field of financial data science by demonstrating the effectiveness of NLP-driven sentiment analysis and offering a framework for building predictive systems that combine textual and numerical financial indicators.
]]> This project aims to predict the statistical performance of college basketball players upon entering the NBA by integrating structured data (rookie-year statistics) with unstructured data (text-based scouting reports). The proposed system applies NLP techniques to extract features from scouting narratives, focusing on player strengths, weaknesses, and comparisons. These textual features are combined with quantitative performance metrics to develop predictive models for key statistical categories such as points, assists, rebounds, blocks, and steals.
Clustering methods are further applied to group players with similar scouting profiles. Instead of labeling clusters as fixed archetypes, the approach emphasizes identifying the most comparable NBA players for each prospect. This similarity-based framework provides interpretability for scouting evaluations and a practical foundation for projecting performance. The predictive component then leverages historical outcomes of comparable players to generate forecasts for new prospects.


The results demonstrate that clustering can organize players into meaningful groups based on textual analysis, while the predictive framework produces reasonable estimates of rookie-year performance. Together, these findings highlight the potential of integrating NLP and ML into scouting workflows to complement traditional evaluation methods, reduce bias, and enhance the accuracy of performance projections.
]]> ]]> ]]> ]]> ]]> The analyzer has gone through several updates: CRAv2 set the foundation with Random Forest classifiers, CRAv3 took a semantic-first approach, and CRAv4 introduced a mixed strategy that fuses rules, semantic pattern analysis, and weighted confidence scoring. There was also an experimental Smart Repository Classifier (SRC) that tried to combine rule-based and machine-learning techniques, plus a side project that looked into detecting design patterns using AST embeddings and Code Property Graphs.
The system was trained on 122 repositories and tested on 47 others across seven different categories. The results were impressive: CRAv3 hit an overall accuracy of 26.4%, while CRAv4 shot up to 70%, with big improvements in web applications (+133%) and CLI tools (+45%). Even though Random Forest tests only maxed out at about 35% in some categories, CRAv4's hybrid method turned out to be both more accurate and easier to understand.
Some key hurdles included the lack of comprehensive datasets, issues with ASTMiner's dependencies, and the author's initial unfamiliarity with machine learning. Tackling these challenges taught valuable lessons in building datasets, feature engineering, and evaluating model

In summary, this project highlights the benefits of blending semantic, structural, and rule-based strategies to gauge developer skills. Looking ahead, plans involve expanding datasets, diving into deep learning for design pattern detection, and enhancing the analyzer into a more general code intelligence tool.
]]>