A comprehensive index of artificial intelligence and machine-learning research with AI-generated summaries, citation metrics, and direct links to papers and code.
Unknown
This paper proposes a Graph RAG-based fault diagnosis system for train bogies that integrates knowledge graphs with large language models to enable dynamic knowledge extraction and improved diagnostic accuracy.
Boyang Zhang, Adrian Lyjak, Eli Stewart, et al.
ExtractBench is a benchmark for schema-guided enterprise document extraction, evaluating value accuracy, record completeness, grounding, and cost across 370 documents.
Unknown
MemReader introduces a reasoning-driven, selective memory extraction approach for long-term agent memory, reducing noise and improving performance.
Liang Hong, Xiao Sun, Yunlei Sun, et al.
This review outlines deep learning methods for automatic text feature extraction, contrasting them with traditional handcrafted approaches.
Maryam M. Najafabadi, Flavio Villanustre, Taghi M. Khoshgoftaar, et al.
Explores how deep learning addresses big data analytics challenges like pattern extraction, semantic indexing, and scalability.
E. Bertin, S. Arnouts
SExtractor is a software tool for automated detection, deblending, measurement, and classification of astronomical sources using neural networks for star/galaxy separation.
Kefallinos, Dionysios, Alexandris, Georgios, Maras, Alexis, et al.
Proposes a framework for answering qualitative spatial questions using a geoparser, crisp reasoner, and answer extraction, with promising synthetic evaluation.
Unknown
NuExtract is a family of small language models fine-tuned on synthetic data to perform structured text-to-JSON extraction, rivaling GPT-4 in accuracy while being orders of magnitude smaller.
Unknown
LMDX adapts arbitrary LLMs for document information extraction with layout encoding and grounding to prevent hallucinations.
Unknown
Introduces Document Screenshot Embedding (DSE), a retrieval method that encodes document screenshots directly into dense vectors using a vision-language model, bypassing content extraction.
Unknown
CEMRAG integrates interpretable visual concept extraction with multimodal RAG to improve both interpretability and factual accuracy in radiology report generation.
F. J. Bolaños, Angelo A. Salatino, Francesco Osborne, et al.
This paper reviews AI techniques for semi-automating systematic literature reviews, focusing on screening and extraction phases.