no code implementations • 15 Dec 2023 • Yuxin Zi, Hariram Veeramani, Kaushik Roy, Amit Sheth
Natural language understanding (NLU) using neural network pipelines often requires additional context that is not solely present in the input data.
no code implementations • 24 Jun 2023 • Yuxin Zi, Kaushik Roy, Vignesh Narayanan, Manas Gaur, Amit Sheth
Crowdsourced and expert-curated knowledge graphs such as ConceptNet are designed to capture the meaning of words from a compact set of well-defined contexts.
no code implementations • 23 Jun 2023 • Kaushik Roy, Yuxin Zi, Vignesh Narayanan, Manas Gaur, Amit Sheth
However, the ad-hoc nature of existing methods makes it difficult to properly analyze the effects of knowledge infusion on the many moving parts or components of a transformer.
no code implementations • 16 Jun 2023 • Kaushik Roy, Yuxin Zi, Manas Gaur, Jinendra Malekar, Qi Zhang, Vignesh Narayanan, Amit Sheth
In this study, we introduce Process Knowledge-infused Learning (PK-iL), a new learning paradigm that layers clinical process knowledge structures on language model outputs, enabling clinician-friendly explanations of the underlying language model predictions.
Explainable Artificial Intelligence (XAI) Language Modelling
1 code implementation • 1 Jun 2023 • Revathy Venkataramanan, Kaushik Roy, Kanak Raj, Renjith Prasad, Yuxin Zi, Vignesh Narayanan, Amit Sheth
In this study, we explore 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.).
no code implementations • 8 May 2023 • Kaushik Roy, Tarun Garg, Vedant Palit, Yuxin Zi, Vignesh Narayanan, Amit Sheth
However, they do not ascribe object and concept-level meaning and semantics to the learned stochastic patterns such as those described in knowledge graphs.
no code implementations • 9 Oct 2022 • Kaushik Roy, Yuxin Zi, Vignesh Narayanan, Manas Gaur, Amit Sheth
Domain-specific language understanding requires integrating multiple pieces of relevant contextual information.