Supplementary Repository: Engineering Control in LLM-Based Recommender Systems (2020–2025)

DescriptionThis supplementary repository centralizes the traceability and reproducibility artifacts associated with the study on LLM-based Conversational Recommender Systems (2020–2025). It provides a single access point for the main materials referenced throughout the manuscript, including screening records, curated datasets, review support artifacts, and extracted textual evidence from the selected studies.The purpose of this repository is to avoid distributing these materials across multiple references in the paper and instead provide one centralized record for inspection, verification, and reuse.Related GitHub repositoryhttps://github.com/JessusTM/Archssistant.gitTable of contentsreview_prompt.txt: Prompt used in an AI-assisted agent during the final full-text review step to support study assessment.CRS - SMS - FINAL.csv: Final consolidated dataset produced from the study and used for the reported analysis.CRS - SMS - Identified Components.csv: Extracted textual evidence from the selected papers indicating, explicitly in the original text and with citation traceability, the number of identified components reported in each study.CRS - SMS - Stage 1 - IC1 IC2 IC3 - EC1 EC2 EC3 EC4.csv: Results of the first screening stage, applying the initial inclusion and exclusion criteria.CRS - SMS - Stage 2 - IC4 EC5 EC6.csv: Results of the second screening stage, applying additional eligibility criteria in the review workflow.CRS - SMS - Stage 3 - EC7.csv: Results of the third screening stage, corresponding to the full-text assessment step and its associated exclusion criterion.CRS - SMS - Stage 4 - IC5.csv: Results of the final screening stage, applying the last inclusion criterion to determine the final set of selected studies.CRS - SMS - CHALLENGES.csv: Extracted textual evidence from the selected papers reporting the challenges identified in each study, preserved with citation traceability.Study Case - Monolithic.pdf: Illustrative study case showing how the system operates in a monolithic architecture scenario, including example input and output together with prototype screenshots.Study Cases - Microservices.pdf: Illustrative study case showing how the system operates in a microservices architecture scenario, including example input and output together with prototype screenshots.A Systematic Mapping Study for LLM-Based Recommender Systems: Full paper manuscript of the systematic mapping study, including protocol, study selection, classification scheme, and main results.This repository is provided to support transparency, traceability, reproducibility, and secondary analyses. No redistributed full-text PDFs of the primary studies are included in this record.