Applied NLP and ML

Odinson

Research on executable, editable extraction programs that combine linguistic structure, neural assistance, provenance, and human review.

AgScribe

LLM-driven structured information extraction over transcribed interviews, using private model deployment and an explicit target schema.

Daedalus

Vision-based PI&D understanding for engineering PDFs, combining document analysis, extraction, and graph-oriented structure.

Corpus Syntax and AI

From Symbolic to Neuro-Symbolic Information Extraction

Active Learning Design Choices for NER with Transformers

METHOD AND SYSTEM FOR CONVERTING LITERATURE INTO A DIRECTED GRAPH

Uncovering linguistic features of intergroup bias in open-ended descriptions of synthesized human images

From Knowledge Discovery to Knowledge Creation: How can Literature-based Discovery Accelerate Progress in Science?

Text to Causal Knowledge Graph: A Framework to Synthesize Knowledge from Unstructured Business Texts into Causal Graphs

A Human-machine Interface for Few-shot Rule Synthesis for Information Extraction

System demonstration for example-driven synthesis of transparent information extraction rules.

From Examples to Rules: Neural Guided Rule Synthesis for Information Extraction

Neural-Guided Program Synthesis of Information Extraction Rules Using Self-Supervision

Syntax-driven Data Augmentation for Named Entity Recognition

Country-level Arabic dialect identification using RNNs with and without linguistic features

Community-guided Hypothesis Generation

Exploring Interpretability in Event Extraction: Multitask Learning of a Neural Event Classifier and an Explanation Decoder

Generating scientific hypotheses through machine reading

Odinson: A Fast Rule-based Information Extraction Framework

Introduces the Odinson rule-based information extraction framework and its fast indexed runtime.

Bridging Non-interacting Research Communities Through Machine-guided Discovery Synthesis

Enabling Search and Collaborative Assembly of Causal Interactions Extracted from Multilingual and Multi-domain Free Text

Demonstrates multilingual causal interaction extraction and collaborative assembly of causal models.

Large-scale Automated Machine Reading Discovers New Cancer Driving Mechanisms

Presents Reach as a large-scale biomedical machine reading system for mechanistic cancer biology.

Machine Reading for Scientific Discovery

Dissertation on machine reading methods for extracting claims from scientific literature, assembling models, and synthesizing hypotheses.

Scientific Discovery as Link Prediction in Influence and Citation Graphs

Swanson linking revisited: Accelerating literature-based discovery across domains using a conceptual influence graph

System demonstration for literature-based discovery with conceptual influence graphs.

Odin's Runes: A Rule Language for Information Extraction

SnapToGrid: From Statistical to Interpretable Models for Biomedical Information Extraction

A Domain-independent Rule-based Framework for Event Extraction