An Empirically Validated Framework for Automated and Personalized Residential Energy-Management Integrating Large Language Models and the Internet of Energy
Round 1
Reviewer 1 Report
Comments and Suggestions for AuthorsThe manuscript presents an innovative residential energy management system that integrates IoT sensor networks with LLMs. The main objective of the work is to provide personalized recommendations for reducing electricity consumption through an accessible conversational interface adapted to different socioeconomic profiles. The research design is clear and presents considerable contributions to the advancement of residential energy management with smart solutions. However, to improve the readability of the work, the following considerations should be addressed.
1 – The introductory section is well-structured and effectively contextualizes the research objective in detail. I recommend including the description of the sections that follow at the end of this section, as a way of introducing the next steps.
2 – In lines 258 to 260 it was mentioned, “The study employed a 12-month longitudinal design, comprising an initial 3-month phase for baseline data collection (without intervention), followed by 9 months of active implementation of the MELISSA system.”. What do the authors mean by “without intervention”? This could lead to the understanding that the data were not processed in any way, or something similar. Please clarify this in the text, as the meaning of this word leads to many interpretations.
3 – This study employed several computational models, including k-means, ARIMA, and VIF multicollinearity analysis. However, in which programming language were they implemented? What are the machine requirements needed to reproduce this study? This information is necessary to improve understanding and enable the reproduction of the methodology.
4 - Which components of the MELISSA system (IoT data collection, statistical analysis, LLM interface) are the most decisive for the effectiveness in reducing energy consumption? I suggest highlighting this in the discussions of this study.
5 – Knowing the importance of generalizing models and studies found in the literature, how can the MELISSA system be adapted to different cultural, climatic, and regulatory contexts outside Brazil? I recommend including this contribution in the manuscript.
6 – It was identified that the acronyms IoT and LLM were described in full more than once. I request that you revise the text to present the description only once and then use the abbreviation thereafter.
The work is well-written and structured. The suggestions made are intended to enrich this work further.
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Reviewer 2 Report
Comments and Suggestions for AuthorsMy review comments are as follows:
- This paper primarily summarizes the advantages of MELISSA and how it leads to an increase in energy efficiency. However, it is lack of technical information about the MELISSA structure and how LLM is integrated with MELISSA.
- The authors could give into more detail about the machine learning techniques that MELISSA uses, particularly with regard to model validation, training, and selection criteria. The work would be much strengthened by further technical information or supplemental materials that offer repeatable pseudocode or data examples.
- Could you please describe how the data analysis agent operates, how the data is examined, and how the Energy management agent is used to provide recommendations?
- Of the 100 houses that take part in the survey, only 97 are analyzed. Could you elaborate on the cause? How do the other three households fare?
- From the paper I saw the benefit of MELISSA framework, while how the framework algorithms works is not explained in detailed. Please explain how this framework save the household energy and lead to meaningful energy consumption reduction.
- I am aware that the outcome demonstrates that "MELISSA" encourages long-term awareness and the ongoing adoption of energy-saving strategies in addition to changing consumption behavior in the short term. Could you elaborate on how this operates, particularly the MELISSA built-in algorithm that alters consumption patterns and raises awareness?
- I would recommend put the section 4.4 at the beginning of the paper. Because it is important to explain what is inside and the technical details of MELISSA engine.
- There is some label overlap and format issue with Figure 7.
- The section discussing security and privacy briefly mentions anonymization and encryption but lacks explicit details or references to standards/protocols adopted (e.g., GDPR compliance).
Author Response
Please see the attachment.
Author Response File: Author Response.pdf
Round 2
Reviewer 1 Report
Comments and Suggestions for AuthorsI thank the authors for answering all the questions. I have no further questions.
Reviewer 2 Report
Comments and Suggestions for AuthorsThank you to the authors for their effort in revising the manuscript and carefully addressing my review comments. After reviewing the revised version, I confirm that all of my previous concerns have been satisfactorily addressed.