A Survey on Event Extraction for Natural Language Understanding: Riding the Biomedical Literature Wave

Luca

Dardari

et al. 2022

Sensors

Self Cite

This paper studies the problem of detecting human beings in non-line-of-sight (NLOS) conditions using an ultra-wideband radar. We perform an extensive measurement campaign in realistic environments, considering different body orientations, the obstacles’ materials, and radar–obstacle distances. We examine two main scenarios according to the radar position: (i) placed on top of a mobile cart; (ii) handheld at different heights. We empirically analyze and compare several input representations and machine learning (ML) methods—supervised and unsupervised, symbolic and non-symbolic—according to both their accuracy in detecting NLOS human beings and their adaptability to unseen cases. Our study proves the effectiveness and flexibility of modern ML techniques, avoiding environment-specific configurations and benefiting from knowledge transference. Unlike traditional TLC approaches, ML allows for generalization, overcoming limits due to unknown or only partially known observation models and insufficient labeled data, which usually occur in emergencies or in the presence of time/cost constraints.

Section: Discussionmentioning

confidence: 93%

Human Being Detection from UWB NLOS Signals: Accuracy and Generality of Advanced Machine Learning Models

Luca

Dardari

et al. 2022

Sensors

Self Cite

“…Moreover, we proved that the method generates semantically accurate summaries in legal datasets, hence it can be successfully applied to other less complex domains. We envisage further directions to deal with inputs longer than the GPU memory allows: i) training models to selfannotate cross-chunk salient information through memorybased neural layers (Moro et al 2018;Cui and Hu 2021); ii) extracting relevant texts with term weighting techniques (Domeniconi et al 2015b) and inter-chunk semantic relations with unsupervised methods (Domeniconi et al 2014a(Domeniconi et al , 2016b to model interpretable representations with knowledge graph (Frisoni and Moro 2021;Frisoni, Moro, and Car-5 https://www.unibo.it/sitoweb/gianluca.moro/en bonaro 2020) or relation and event extraction (Domeniconi et al 2016a;Frisoni, Moro, and Carbonaro 2021).…”

Section: Discussionmentioning

confidence: 99%

Semantic Self-Segmentation for Abstractive Summarization of Long Documents in Low-Resource Regimes

Ragazzi

2022

AAAI

Self Cite

The quadratic memory complexity of transformers prevents long document summarization in low computational resource scenarios. State-of-the-art models need to apply input truncation, thus discarding and ignoring potential summary-relevant contents, leading to a performance drop. Furthermore, this loss is generally destructive for semantic text analytics in high-impact domains such as the legal one. In this paper, we propose a novel semantic self-segmentation (Se3) approach for long document summarization to address the critical problems of low-resource regimes, namely to process inputs longer than the GPU memory capacity and produce accurate summaries despite the availability of only a few dozens of training instances. Se3 segments a long input into semantically coherent chunks, allowing transformers to summarize very long documents without truncation by summarizing each chunk and concatenating the results. Experimental outcomes show the approach significantly improves the performance of abstractive summarization transformers, even with just a dozen of labeled data, achieving new state-of-the-art results on two legal datasets of different domains and contents. Finally, we report ablation studies to evaluate each contribution of the components of our method to the performance gain.

“…Events are key for revealing underlying real-world knowledge and capturing biological processes described in the unstructured text. Unfortunately, many contributions on event extraction [ 2 ] are contrasted by a tiny number of works on event representation learning, which constitutes a pressing need. Event embedding is an efficient solution to represent discrete and sparse events as dense vectors in a continuous space, reflecting their similarity and availing numerous applications.…”

Section: Related Workmentioning

confidence: 99%

“…According to bibliometric results verifiable at https://pubmed.ncbi.nlm.nih.gov/ using the query "1700:2021[dp]"-accessed on 11 December 2021-the annual rate of biomedical publications registered on PubMed is growing exponentially. As biology literature increases in the form of text documents, the automatic extraction of biomedical entities (e.g., proteins, genes, diseases, drugs) and their semantic relations have been intensively investigated [2]. Recent developments in deep learning (DL) and natural language processing (NLP) have enabled intelligent ways to uncover structured, concise, and unambiguous knowledge mentioned in large unstructured text corpora.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Unsupervised Event Graph Representation and Similarity Learning on Biomedical Literature

Frisoni

Carlassare

et al. 2021

Sensors

Self Cite

The automatic extraction of biomedical events from the scientific literature has drawn keen interest in the last several years, recognizing complex and semantically rich graphical interactions otherwise buried in texts. However, very few works revolve around learning embeddings or similarity metrics for event graphs. This gap leaves biological relations unlinked and prevents the application of machine learning techniques to promote discoveries. Taking advantage of recent deep graph kernel solutions and pre-trained language models, we propose Deep Divergence Event Graph Kernels (DDEGK), an unsupervised inductive method to map events into low-dimensional vectors, preserving their structural and semantic similarities. Unlike most other systems, DDEGK operates at a graph level and does not require task-specific labels, feature engineering, or known correspondences between nodes. To this end, our solution compares events against a small set of anchor ones, trains cross-graph attention networks for drawing pairwise alignments (bolstering interpretability), and employs transformer-based models to encode continuous attributes. Extensive experiments have been done on nine biomedical datasets. We show that our learned event representations can be effectively employed in tasks such as graph classification, clustering, and visualization, also facilitating downstream semantic textual similarity. Empirical results demonstrate that DDEGK significantly outperforms other state-of-the-art methods.