|
Translating the Pulse of Poe: A Stylistic Comparison of Rhythm in English and Romanian Versions Florin-Cristian Motrun, Ştefan Trăuşan-Matu |
65-84 |
|
Real Estate Transactions System Enhanced with Machine Learning Algorithms Andreea Bianca Popa, Paul Stefan Popescu |
85-96 |
1 National University of Science and Technology Politehnica Bucharest
313 Splaiul Independentei, Bucharest, Romania
2 Institutul de Cercetări în Inteligenţa Artificială
Calea 13 Septembrie nr. 13, Bucureşti
3 Academy of Romanian Scientists
Splaiul Indpendentei 54, Bucharest, Romania
Abstract. The paper analyzes the rhythmic profile of Edgar Allan Poe's poems together with Emil Gulian's Romanian translations. It is presented as a pipeline that integrates three Romanian mathematical models developed by Solomon Marcus, Mihai Dinu, and Vasile Vasile, with contemporary natural language processing methods to automate syllable extraction, stress assignment, and rhythmic pattern detection. Repetition phenomena and additional rhythmic devices are also quantified. Applying the system to aligned poem pairs reveals how Poe's original rhythmic structures are preserved, amplified, or reshaped in translation, providing complementary quantitative metrics and qualitative insights into cross‑lingual poetic rhythm.
Keywords: rhythm analysis, rhythmic devices, stress, syllable, natural language processing, translation
Cite this paper as:
Motrun, F.-C., Trăuşan-Matu, S. Translating the Pulse of Poe: A Stylistic Comparison of Rhythm in English and Romanian Versions.
International Journal of User-System Interaction 17(3),
65-84, 2024.
University of Craiova
Craiova, Romania
Abstract. Real-estate transactions are interaction-intensive processes that require property discovery, negotiation, and frequent coordination between clients and agents. This paper presents the design and implementation of a web platform that supports end-to-end user interaction around residential listings: role-based authentication (client/agent/admin), advanced search and filtering, wishlist management, internal property-contextual messaging, and meeting requests for on-site visits. To support decision-making, the platform integrates automated price estimation using supervised machine learning. The prediction module is deployed as a Python (Flask) microservice and is consumed by the PHP backend through a REST-style interface, enabling modular evolution of the intelligent component. Several regression algorithms were evaluated using a real-world dataset, and Random Forest achieved the best performance (MAE 41,634.65; RMSE 86,272.47; R² 0.71). The resulting system demonstrates how interaction features and data-driven intelligence can be integrated into a cohesive, practical solution for digital real estate workflows.
Keywords: user-system interaction; real-estate platform; web application
Cite this paper as:
Popa, A. B., Popescu, P. S. Real Estate Transactions System Enhanced with Machine Learning Algorithms.
International Journal of User-System Interaction 17(3),
85-96, 2024.