Revathy Venkataramanan

Revathy Venkataramanan

Auto-generated profile, based on publicly available information

Revathy Venkataramanan

University of South Carolina, Columbia, SC, USA

Research Focus & Metrics

Academic Career Length

10Years (2017-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
14 December 2018 in JMIR Pediatrics and Parenting
Verified
Journal Article
Open access
UJUtkarshani Jaimini
Utkarshani Jaimini
KTKrishnaprasad Thirunarayan
Krishnaprasad Thirunarayan
MKManinder Kalra
Maninder Kalra
RVRevathy Venkataraman
Revathy Venkataraman
DKDipesh Kadariya
Dipesh Kadariya
Amit ShethAmit Sheth
Amit Sheth
In the traditional asthma management protocol, a child meets with a clinician infrequently, once in 3 to 6 months, and is assessed using the Asthma Control Test questionnaire. This information is inadequate for timely determination of asthma control, compliance, precise diagnosis of the cause, and assessing the effectiveness of the treatment plan. The continuous monitoring and improved tracking of the child’s symptoms, activities, sleep, and treatment adherence can allow precise determination of asthma triggers and a reliable assessment of medication compliance and effectiveness. Digital phenotyping refers to moment-by-moment quantification of the individual-level human phenotype in situ using data from personal digital devices, in particular, mobile phones. The kHealth kit consists of a mobile app, provided on an Android tablet, that asks timely and contextually relevant questions related to asthma symptoms, medication intake, reduced activity because of symptoms, and nighttime awakenings; a Fitbit to monitor activity and sleep; a Microlife Peak Flow Meter to monitor the peak expiratory flow and forced exhaled volume in 1 second; and a Foobot to monitor indoor air quality. The kHealth cloud stores personal health data and environmental data collected using Web services. The kHealth Dashboard interactively visualizes the collected data. The objective of this study was to discuss the usability and feasibility of collecting clinically relevant data to help clinicians diagnose or intervene in a child’s care plan by using the kHealth system for continuous and comprehensive monitoring of child’s symptoms, activity, sleep pattern, environmental triggers, and compliance. The kHealth system helps in deriving actionable insights to help manage asthma at both the personal and cohort levels. The Digital Phenotype Score and Controller Compliance Score introduced in the study are the basis of ongoing work on addressing personalized asthma care and answer questions such as, “How can I help my child better adhere to care instructions and reduce future exacerbation?” The Digital Phenotype Score and Controller Compliance Score summarize the child’s condition from the data collected using the kHealth kit to provide actionable insights. The Digital Phenotype Score formalizes the asthma control level using data about symptoms, rescue medication usage, activity level, and sleep pattern. The Compliance Score captures how well the child is complying with the treatment protocol. We monitored and analyzed data for 95 children, each recruited for a 1- or 3-month-long study. The Asthma Control Test scores obtained from the medical records of 57 children were used to validate the asthma control levels calculated using the Digital Phenotype Scores. At the cohort level, we found asthma was very poorly controlled in 37% (30/82) of the children, not well controlled in 26% (21/82), and well controlled in 38% (31/82). Among the very poorly controlled children (n=30), we found 30% (9/30) were highly compliant toward their controller medication intake—suggesting a re-evaluation for change in medication or dosage—whereas 50% (15/30) were poorly compliant and candidates for a more timely intervention to improve compliance to mitigate their situation. We observed a negative Kendall Tau correlation between Asthma Control Test scores and Digital Phenotype Score as –0.509 (P<.01). kHealth kit is suitable for the collection of clinically relevant information from pediatric patients. Furthermore, Digital Phenotype Score and Controller Compliance Score, computed based on the continuous digital monitoring, provide the clinician with timely and detailed evidence of a child’s asthma-related condition when compared with the Asthma Control Test scores taken infrequently during clinic visits.

27 June 2019 in JMIR Pediatrics and Parenting
Verified
Journal Article
Open access
RVRevathy Venkataramanan
Revathy Venkataramanan
KTKrishnaprasad Thirunarayan
Krishnaprasad Thirunarayan
UJUtkarshani Jaimini
Utkarshani Jaimini
DKDipesh Kadariya
Dipesh Kadariya
HYHong Yung Yip
Hong Yung Yip
Amit ShethAmit Sheth
Amit Sheth
Background: Asthma is a chronic pulmonary disease with multiple triggers. It can be managed by strict adherence to an asthma care plan and by avoiding these triggers. Clinicians cannot continuously monitor their patients’ environment and their adherence to an asthma care plan, which poses a significant challenge for asthma management. Objective: In this study, pediatric patients were continuously monitored using low-cost sensors to collect asthma-relevant information. The objective of this study was to assess whether kHealth kit, which contains low-cost sensors, can identify personalized triggers and provide actionable insights to clinicians for the development of a tailored asthma care plan. Methods: The kHealth asthma kit was developed to continuously track the symptoms of asthma in pediatric patients and monitor the patients’ environment and adherence to their care plan for either 1 or 3 months. The kit consists of an Android app–based questionnaire to collect information on asthma symptoms and medication intake, Fitbit to track sleep and activity, the Peak Flow meter to monitor lung functions, and Foobot to monitor indoor air quality. The data on the patient’s outdoor environment were collected using third-party Web services based on the patient’s zip code. To date, 107 patients consented to participate in the study and were recruited from the Dayton Children’s Hospital, of which 83 patients completed the study as instructed. Results: Patient-generated health data from the 83 patients who completed the study were included in the cohort-level analysis. Of the 19% (16/83) of patients deployed in spring, the symptoms of 63% (10/16) and 19% (3/16) of patients suggested pollen and particulate matter (PM2.5), respectively, to be their major asthma triggers. Of the 17% (14/83) of patients deployed in fall, symptoms of 29% (4/17) and 21% (3/17) of patients suggested pollen and PM2.5, respectively, to be their major triggers. Among the 28% (23/83) of patients deployed in winter, PM2.5 was identified as the major trigger for 83% (19/23) of patients. Similar correlations were not observed between asthma symptoms and factors such as ozone level, temperature, and humidity. Furthermore, 1 patient from each season was chosen to explain, in detail, his or her personalized triggers by observing temporal associations between triggers and asthma symptoms gathered using the kHealth asthma kit. Conclusions: The continuous monitoring of pediatric asthma patients using the kHealth asthma kit generates insights on the relationship between their asthma symptoms and triggers across different seasons. This can ultimately inform personalized asthma management and intervention plans.

1 October 2023
Verified
Conference Paper
Open access
RVRevathy Venkataramanan
Revathy Venkataramanan
Kaushik RoyKaushik Roy
Kaushik Roy
KRKanak Raj
Kanak Raj
RPRenjith Prasad
Renjith Prasad
YZYuxin Zi
Yuxin Zi
ASAmit Sheth
Amit Sheth
As people become more aware of their food choices, food computation models have become increasingly popular in assisting people in maintaining healthy eating habits. For example, food recommendation systems analyze recipe instructions to assess nutritional contents and provide recipe recommendations. The recent and remarkable successes of generative AI methods, such as auto-regressive Large Language Models, can enable robust methods for a more comprehensive understanding of recipes for healthy food recommendations beyond surface-level nutrition content assessments. In this study, we investigate the use of generative AI methods to extend current food computation models, primarily involving the analysis of nutrition and ingredients, to also incorporate cooking actions (e.g., add salt, fry the meat, boil the vegetables, etc.), Cooking actions are notoriously hard to model using statistical learning methods due to irregular data patterns - significantly varying natural language descriptions for the same action (e.g., marinate the meat vs. marinate the meat and leave overnight) and infrequently occurring patterns (e.g., add salt occurs far more frequently than marinating the meat). The prototypical approach to handling irregular data patterns is to increase the volume of data that the model ingests by orders of magnitude. Unfortunately, in the cooking domain, these problems are further compounded with larger data volumes presenting a unique challenge that is not easily handled by simply scaling up. In this work, we propose novel aggregation-based generative AI methods, Cook-Gen, that reliably generate cooking actions from recipes, despite difficulties with irregular data patterns, while also outperforming Large Language Models and other strong baselines.

4 December 2024 in IEEE Intelligent Systems
Verified
Journal Article
Closed access
DTDeepa Tilwani
Deepa Tilwani
RVRevathy Venkataramanan
Revathy Venkataramanan
Amit ShethAmit Sheth
Amit Sheth
Attribution in large language models (LLMs) remains a significant challenge, particularly in ensuring the factual accuracy and reliability of the generated outputs. Current methods for citation or attribution, such as those employed by tools like Perplexity.ai and Bing-Search-integrated LLMs, attempt to ground responses by providing real-time search results and citations. However, so far, these approaches suffer from issues such as hallucinations; biases; surface-level relevance matching; and the complexity of managing vast, unfiltered knowledge sources. While tools like Perplexity.ai dynamically integrate Web-based information and citations, they often rely on inconsistent sources, such as blog posts or unreliable sources, which limits their overall reliability. We explain that these challenges can be mitigated by integrating neurosymbolic AI (NesyAI), which combines the strengths of neural networks, with structured symbolic reasoning. NesyAI offers transparent, interpretable, and dynamic reasoning processes, addressing the limitations of current attribution methods by incorporating structured symbolic knowledge with flexible, neural-based learning. This article explores how NesyAI frameworks can enhance existing attribution models, offering more reliable, interpretable, and adaptable systems for LLMs.

17 March 2025 in IEEE Internet Computing
Verified
Journal Article
Closed access
RVRevathy Venkataramanan
Revathy Venkataramanan
CKChathurangi Shyalika Jayakody Kankanamalage
Chathurangi Shyalika Jayakody Kankanamalage
Amit ShethAmit Sheth
Amit Sheth
Compositional reasoning, the cognitive process of breaking complex problems into manageable subproblems and recomposing them to generate new ideas, is fundamental to human problem solving and critical thinking. While deep learning models excel at pattern recognition, their capacity for true understanding and reasoning remains a topic of debate. Although the growth of the Internet has provided the necessary scale of data for model training, the way data is represented plays a pivotal role in enabling reasoning capabilities. This article introduces dynamic multimodal process knowledge graphs (DMPKGs), a novel neurosymbolic framework for data and knowledge representation that supports cognitive tasks such as compositional reasoning, high-level abstraction, explainability, and causal inference along with representation learning. The framework integrates data and knowledge into a unified, structured format enriched with semantics from multiple contexts. By prioritizing contextualized and semantically rich representations, DMPKGs aim to bridge the gap between pattern recognition and reasoning in artificial intelligence systems.

4 September 2025 in IEEE Internet Computing
Verified
Journal Article
Closed access
RGRitvik Garimella
Ritvik Garimella
HYHong Yung Yip
Hong Yung Yip
RVRevathy Venkataramanan
Revathy Venkataramanan
Amit ShethAmit Sheth
Amit Sheth
As enterprises increasingly aim to incorporate artificial intelligence into their workflows to tackle complex, multimodal tasks, the demand for intelligent, robust, and trustworthy systems is paramount. While multimodal large language models offer initial capabilities for processing diverse data streams, their dependence on embedding-based representations limit their effectiveness in delivering semantically grounded explanations and reasoning, as well as qualities essential for enterprise-grade applications. Neurosymbolic approaches provide a promising alternative by enabling traceable, context-aware decision making. However, constructing enterprise-level multimodal knowledge graphs (MMKGs) that enable neurosymbolic approaches remains largely impractical. Although prior efforts have explored MMKG construction, they fall short in addressing the scalability, modularity, and integration that are necessary for any enterprise-grade application. We present a fully automated MMKG construction framework tailored to real-world enterprise environments. Our system features a modular, self-refining lifecycle with a support for human-in-the-loop feedback, enabling scalable, cost-effective, and task-aligned MMKG generation. We demonstrate the practical value of our framework through a real-world case study, showcasing its ability to transform unstructured multimodal data into actionable, semantically grounded knowledge assets for enterprise use.