• Publications
  • Influence
Learning from Sparse Data by Exploiting Monotonicity Constraints
TLDR
This paper shows how to interpret knowledge of qualitative influences, and in particular of monotonicities, as constraints on probability distributions, and to incorporate this knowledge into Bayesian network learning algorithms.
Learning first-order probabilistic models with combining rules
TLDR
A language that consists of quantified conditional influence statements and captures most relational probabilistic models based on directed graphs is described and algorithms based on gradient descent and expectation maximization for different combining rules are derived and implemented.
Preference Elicitation via Theory Refinement
We present an approach to elicitation of user preference models in which assumptions can be used to guide but not constrain the elicitation process. We demonstrate that when domain knowledge is
Detecting experimental techniques and selecting relevant documents for protein-protein interactions from biomedical literature
TLDR
A novel approach that converts the multi-class, multi-label classification problem to a binary classification problem showed much promise in IMT, and concluded that contextual words surrounding named entities, as well as the MeSH headings associated with the documents were among the main contributors to the performance.
Text mining for efficient search and assisted creation of clinical trials
TLDR
ASCOT uses text mining and data mining methods to enrich clinical trials with metadata, that in turn serve as effective tools to narrow down search.
ARGUER: using argument schemas for argument detection and rebuttal in dialogs
TLDR
The method can be used to implement a computational agent that is both able to detect arguments and to generate candidate arguments for rebuttal.
Building intelligent dialog systems
TLDR
This work specied and built systems for giving medical students an opportunity to practice their decision making skills in English (B2); performing template-based natural language generation (YAG); detecting and rebutting arguments (ARGUER); recognizing and repairing misunderstandings (RRM); and assessing and augmenting patients' health knowledge (PEAS).
Inferring appropriate eligibility criteria in clinical trial protocols without labeled data
TLDR
The user task of designing clinical trial protocols is considered and a method that outputs the most appropriate eligibility criteria from a potentially huge set of candidates is proposed, which is 8 and 9 times better than randomly choosing from a set of candidate obtained from relevant documents.
A method for discovering and inferring appropriate eligibility criteria in clinical trial protocols without labeled data
TLDR
Results from the experiments suggest that LDALR models can be used to effectively find appropriate eligibility criteria from a large repository of clinical trial protocols.
Eliciting Utilities by Refining Theories of Monotonicity and Risk
TLDR
This work uses knowledge-based artificial neural networks to encode assumptions about a decision maker’s preferences, and presents empirical results showing that learning speed and accuracy increase as more domain knowledge is included in the neural net.
...
...