Relation Extraction Research Articles

Improved relation extraction through key phrase identification using community detection on dependency trees

A method for extracting relations from sentences by utilizing their dependency trees to identify key phrases is presented in this paper. Dependency trees are commonly used in natural language processing to represent the grammatical structure of a sentence, and this approach builds upon this representation to extract meaningful relations between phrases. Identifying key phrases is crucial in relation extraction as they often indicate the entities and actions involved in a relation. The method uses community detection algorithms on the dependency tree to identify groups of related words that form key phrases, such as subject-verb-object structures. The experiments on the Semeval-2010 task8 dataset and the TACRED dataset demonstrate that the proposed method outperforms existing baseline methods.

Adversarial Multi-Teacher Distillation for Semi-Supervised Relation Extraction.

The shortage of labeled data has been a long-standing challenge for relation extraction (RE) tasks. Semi-supervised RE (SSRE) is a promising way through annotating unlabeled samples with pseudolabels as additional training data. However, some pseudolabels on unlabeled data might be erroneous and will bring misleading knowledge into SSRE models. For this reason, we propose a novel adversarial multi-teacher distillation (AMTD) framework, which includes multi-teacher knowledge distillation and adversarial training (AT), to capture the knowledge on unlabeled data in a refined way. Specifically, we first develop a general knowledge distillation (KD) technique to learn not only from pseudolabels but also from the class distribution of predictions by different models in existing SSRE methods. To improve the robustness of the model, we further empower the distillation process with a language model-based AT technique. Extensive experimental results on two public datasets demonstrate that our framework significantly promotes the performance of the base SSRE methods.

PANNER: POS-Aware Nested Named Entity Recognition Through Heterogeneous Graph Neural Network

Nested named entity recognition (Nested NER) in knowledge graph (KG) aims at obtaining all meaningful entities, including nested entities for sentences in longer text region. Those obtained entities are to facilitate downstream applications, such as relation extraction, entity resolution, and coreference resolution. This task, however, is challenging not only because of the demand to detect the boundary of the entity but also due to the complexity of those hierarchically nested entities. Since a substantial amount of work has been made to Flat NER (or Nested NER), a few of them can explicitly acquire the position of the entity and utilize the grammatical construction of text. In this work, we propose PANNER, a POS-aware Nested NER model, to solve all the above issues. Specifically, we first construct a heterogeneous graph by introducing the part-of-speech (POS) information of the word. Second, we design a dilated random walk (DRW) algorithm based on a grammatical path to sample a fixed size of neighbors for each node. Third, we aggregate the message from different types of neighbors through an attention mechanism. Finally, we use a bidirectional decoding module to recognize and categorize all the flat and nested entities based on the node embedding in a layer-wise manner. Our extensive experiments show the effectiveness of PANNER in both flat and nested NER.

Joint extraction of Chinese medical entities and relations based on RoBERTa and single-module global pointer

BackgroundMost Chinese joint entity and relation extraction tasks in medicine involve numerous nested entities, overlapping relations, and other challenging extraction issues. In response to these problems, some traditional methods decompose the joint extraction task into multiple steps or multiple modules, resulting in local dependency in the meantime.MethodsTo alleviate this issue, we propose a joint extraction model of Chinese medical entities and relations based on RoBERTa and single-module global pointer, namely RSGP, which formulates joint extraction as a global pointer linking problem. Considering the uniqueness of Chinese language structure, we introduce the RoBERTa-wwm pre-trained language model at the encoding layer to obtain a better embedding representation. Then, we represent the input sentence as a third-order tensor and score each position in the tensor to prepare for the subsequent process of decoding the triples. In the end, we design a novel single-module global pointer decoding approach to alleviate the generation of redundant information. Specifically, we analyze the decoding process of single character entities individually, improving the time and space performance of RSGP to some extent.ResultsIn order to verify the effectiveness of our model in extracting Chinese medical entities and relations, we carry out the experiments on the public dataset, CMeIE. Experimental results show that RSGP performs significantly better on the joint extraction of Chinese medical entities and relations, and achieves state-of-the-art results compared with baseline models.ConclusionThe proposed RSGP can effectively extract entities and relations from Chinese medical texts and help to realize the structure of Chinese medical texts, so as to provide high-quality data support for the construction of Chinese medical knowledge graphs.

Multi-head CRF classifier for biomedical multi-class named entity recognition on Spanish clinical notes.

The identification of medical concepts from clinical narratives has a large interest in the biomedical scientific community due to its importance in treatment improvements or drug development research. Biomedical named entity recognition (NER) in clinical texts is crucial for automated information extraction, facilitating patient record analysis, drug development, and medical research. Traditional approaches often focus on single-class NER tasks, yet recent advancements emphasize the necessity of addressing multi-class scenarios, particularly in complex biomedical domains. This paper proposes a strategy to integrate a multi-head conditional random field (CRF) classifier for multi-class NER in Spanish clinical documents. Our methodology overcomes overlapping entity instances of different types, a common challenge in traditional NER methodologies, by using a multi-head CRF model. This architecture enhances computational efficiency and ensures scalability for multi-class NER tasks, maintaining high performance. By combining four diverse datasets, SympTEMIST, MedProcNER, DisTEMIST, and PharmaCoNER, we expand the scope of NER to encompass five classes: symptoms, procedures, diseases, chemicals, and proteins. To the best of our knowledge, these datasets combined create the largest Spanish multi-class dataset focusing on biomedical entity recognition and linking for clinical notes, which is important to train a biomedical model in Spanish. We also provide entity linking to the multi-lingual Systematized Nomenclature of Medicine Clinical Terms (SNOMED CT) vocabulary, with the eventual goal of performing biomedical relation extraction. Through experimentation and evaluation of Spanish clinical documents, our strategy provides competitive results against single-class NER models. For NER, our system achieves a combined micro-averaged F1-score of 78.73, with clinical mentions normalized to SNOMED CT with an end-to-end F1-score of 54.51. The code to run our system is publicly available at https://github.com/ieeta-pt/Multi-Head-CRF. Database URL: https://github.com/ieeta-pt/Multi-Head-CRF.

Leveraging shortest dependency paths in low-resource biomedical relation extraction

BackgroundBiomedical Relation Extraction (RE) is essential for uncovering complex relationships between biomedical entities within text. However, training RE classifiers is challenging in low-resource biomedical applications with few labeled examples.MethodsWe explore the potential of Shortest Dependency Paths (SDPs) to aid biomedical RE, especially in situations with limited labeled examples. In this study, we suggest various approaches to employ SDPs when creating word and sentence representations under supervised, semi-supervised, and in-context-learning settings.ResultsThrough experiments on three benchmark biomedical text datasets, we find that incorporating SDP-based representations enhances the performance of RE classifiers. The improvement is especially notable when working with small amounts of labeled data.ConclusionSDPs offer valuable insights into the complex sentence structure found in many biomedical text passages. Our study introduces several straightforward techniques that, as demonstrated experimentally, effectively enhance the accuracy of RE classifiers.

Dielectric Ceramics Database Automatically Constructed by Data Mining in the Literature.

Vast published dielectric ceramics literature is a natural database for big-data analysis, discovering structure-property relationships, and property prediction. We constructed a data-mining pipeline based on natural language processing (NLP) to extract property information from about 12,900 published dielectric ceramics articles and normalized more than 20 properties. The micro-F1 scores for sentence classification, named entities recognition, relation extraction (related), and relation extraction (same), are 91.6, 82.4, 91.4, and 88.3%, respectively. We demonstrated the distribution of some essential properties according to the publication years to reveal the tendency. In order to test the reliability of the data extraction, we trained an XGBoost model to predict the dielectric constant and used the SHAP module to interpret the contribution of each feature in order to identify some of the factors that determine the dielectric properties. The result shows that including Q × f in the model can increase the dielectric constant prediction accuracy. Our work can give some hints to experimentalists on their way to improve the performances of cutting-edge materials.

Let’s Discover More API Relations: A Large Language Model-based AI Chain for Unsupervised API Relation Inference

APIs have intricate relations that can be described in text and represented as knowledge graphs to aid software engineering tasks. Existing relation extraction methods have limitations, such as limited API text corpus and affected by the characteristics of the input text. To address these limitations, we propose utilizing large language models (LLMs) (e.g., gpt-3.5) as a neural knowledge base for API relation inference. This approach leverages the entire Web used to pre-train LLMs as a knowledge base and is insensitive to the context and complexity of input texts. To ensure accurate inference, we design an AI chain consisting of three AI modules: API Fully Qualified Name (FQN) Parser, API Knowledge Extractor, and API Relation Decider. The accuracy of the API FQN Parser and API Relation Decider is 0.81 and 0.83, respectively. Using the generative capacity of the LLM and our approach’s inference capability, we achieve an average F1 value of 0.76 under the three datasets, significantly higher than the state-of-the-art method’s average F1 value of 0.40. Compared to the original CoT and modularized CoT methods, our AI chain design has improved the performance of API relation inference by 71% and 49%, respectively. Meanwhile, the prompt ensembling strategy enhances the performance of our approach by 32%. The API relations inferred by our method can be further organized into structured forms to provide support for other software engineering tasks.

A Comprehensive Survey on Relation Extraction: Recent Advances and New Frontiers

Relation extraction (RE) involves identifying the relations between entities from underlying content. RE serves as the foundation for many natural language processing (NLP) and information retrieval applications, such as knowledge graph completion and question answering. In recent years, deep neural networks have dominated the field of RE and made noticeable progress. Subsequently, the large pre-trained language models (PLMs) have taken the state-of-the-art RE to a new level. This survey provides a comprehensive review of existing deep learning techniques for RE. First, we introduce RE resources, including datasets and evaluation metrics. Second, we propose a new taxonomy to categorize existing works from three perspectives, i.e., text representation, context encoding, and triplet prediction. Third, we discuss several important challenges faced by RE and summarize potential techniques to tackle these challenges. Finally, we outline some promising future directions and prospects in this field. This survey is expected to facilitate researchers’ collaborative efforts to address the challenges of real-world RE systems.

TCohPrompt: task-coherent prompt-oriented fine-tuning for relation extraction

Prompt-tuning has emerged as a promising approach for improving the performance of classification tasks by converting them into masked language modeling problems through the insertion of text templates. Despite its considerable success, applying this approach to relation extraction is challenging. Predicting the relation, often expressed as a specific word or phrase between two entities, usually requires creating mappings from these terms to an existing lexicon and introducing extra learnable parameters. This can lead to a decrease in coherence between the pre-training task and fine-tuning. To address this issue, we propose a novel method for prompt-tuning in relation extraction, aiming to enhance the coherence between fine-tuning and pre-training tasks. Specifically, we avoid the need for a suitable relation word by converting the relation into relational semantic keywords, which are representative phrases that encapsulate the essence of the relation. Moreover, we employ a composite loss function that optimizes the model at both token and relation levels. Our approach incorporates the masked language modeling (MLM) loss and the entity pair constraint loss for predicted tokens. For relation level optimization, we use both the cross-entropy loss and TransE. Extensive experimental results on four datasets demonstrate that our method significantly improves performance in relation extraction tasks. The results show an average improvement of approximately 1.6 points in F1 metrics compared to the current state-of-the-art model. Codes are released at https://github.com/12138yx/TCohPrompt.

Kinetic evaluation of Azadirichtin and related phytoconstituents extraction from Azadirachtaindica (neem) and Melia azedarach (chinaberry) seeds using ultrasound assisted extraction and maceration

The comparative extraction of total phenolic compounds from A. indica and M. azedarach seed powder using two different methodologies is evaluated. The solvent screening was done using soxhlet extraction for six different solvents and methanol had the maximum extraction capacity for both seed powder. Extraction of total phenolic compounds from A. indica and M. azedarach seed powder was done using maceration and ultrasound. The kinetics and extraction mechanism was evaluated using popular three empirical and five semi-empirical models along with scanning electron micrograph analysis. The rate constants showed an enhancement in extraction kinetics of about 4.5 fold for M. azedarach and 7.8 fold for A. indica seeds using ultrasound. Based on model evaluation, extraction using maceration for both seeds follows inter-particle diffusion as the rate-limiting step with non-swelling behaviour. Whereas, in the case of ultrasound, the dependence of rate is on the diffusion of solute within the solid and the solid surface to the solvent. In addition, HPLC and GC-MS analysis suggested differences in the composition of extracts and some important phytochemicals were extracted in larger quantities using ultrasound as compared to maceration.

Few-shot relational triple extraction with hierarchical prototype optimization

Relational Triple Extraction (RTE) aims to extract relations and entities from unstructured text. Current RTE models using supervised learning require a large amount of labeled data, which presents a challenge for real-world applications. Therefore, the research work on Few-Shot Relational Triple Extraction (FS-RTE) has been proposed. However, the existing work cannot effectively construct accurate prototypes from a small number of samples, and it is difficult to model the dependencies between entities and relations, resulting in poor performance in relational triple extraction. In this paper, we propose a Hierarchical Prototype Optimized FS-RTE method (HPO). In particular, to mitigate prototype bias built on a small number of samples, HPO uses prompt learning to merge the information of relational labels into the text. Then, the entity-level prototypes are constructed using a span encoder to avoid label dependency between entity tokens. Finally, the hierarchical contrastive learning (HCL) method is introduced to improve the metric space between the prototypes of entities and relations, respectively. Experiments conducted on two public datasets show that HPO can significantly outperform previous state-of-the-art methods.

Joint relational triple extraction with enhanced representation and binary tagging framework in cybersecurity

The cyber threat intelligence (CTI) knowledge graph is a valuable tool for aiding security practitioners in the identification and analysis of cyberattacks. These graphs are constructed from CTI data, organized into relational triples, where each triple comprises two entities linked by a particular relation. However, as the volume of CTI data is expanding at a faster rate than predicted, existing technologies are unable to extract relational triples quickly and accurately. This work mainly focuses on the extraction of relational triples in CTI data, which is achieved by an enhanced representation and binary tagging framework (ERBTF). Firstly, we introduce embedding representations for relations and concatenate these with word embeddings to obtain the initial hidden representation. Subsequently, we employ a novel dilated convolutional encoder that consists of a dilated convolution neural network, gate mechanism and residual connection to enhance the learned contextual representation. Afterwards, we adopt an attention module that includes multi-head self-attention and position-wise feed-forward neural network to allocate greater attention to words that significantly influence the specific relation. Additionally, we utilize the straightforward yet efficient binary entity tagger to identify subject and object entities under different relations for constructing relational triples. We conduct massive experiments on relational triple extraction from CTI data, the results show that ERBTF is superior to the existing relation extraction models, and achieves state-of-the-art performance.

Multi-Level Attention with 2D Table-Filling for Joint Entity-Relation Extraction

Joint entity-relation extraction is a fundamental task in the construction of large-scale knowledge graphs. This task relies not only on the semantics of the text span but also on its intricate connections, including classification and structural details that most previous models overlook. In this paper, we propose the incorporation of this information into the learning process. Specifically, we design a novel two-dimensional word-pair tagging method to define the task of entity and relation extraction. This allows type markers to focus on text tokens, gathering information for their corresponding spans. Additionally, we introduce a multi-level attention neural network to enhance its capacity to perceive structure-aware features. Our experiments show that our approach can overcome the limitations of earlier tagging methods and yield more accurate results. We evaluate our model using three different datasets: SciERC, ADE, and CoNLL04. Our model demonstrates competitive performance compared to the state-of-the-art, surpassing other approaches across the majority of evaluated metrics.

Towards discovery: an end-to-end system for uncovering novel biomedical relations.

Biomedical relation extraction is an ongoing challenge within the natural language processing community. Its application is important for understanding scientific biomedical literature, with many use cases, such as drug discovery, precision medicine, disease diagnosis, treatment optimization and biomedical knowledge graph construction. Therefore, the development of a tool capable of effectively addressing this task holds the potential to improve knowledge discovery by automating the extraction of relations from research manuscripts. The first track in the BioCreative VIII competition extended the scope of this challenge by introducing the detection of novel relations within the literature. This paper describes that our participation system initially focused on jointly extracting and classifying novel relations between biomedical entities. We then describe our subsequent advancement to an end-to-end model. Specifically, we enhanced our initial system by incorporating it into a cascading pipeline that includes a tagger and linker module. This integration enables the comprehensive extraction of relations and classification of their novelty directly from raw text. Our experiments yielded promising results, and our tagger module managed to attain state-of-the-art named entity recognition performance, with a micro F1-score of 90.24, while our end-to-end system achieved a competitive novelty F1-score of 24.59. The code to run our system is publicly available at https://github.com/ieeta-pt/BioNExt. Database URL: https://github.com/ieeta-pt/BioNExt.

Relation Extraction in Underexplored Biomedical Domains: A Diversity-optimised Sampling and Synthetic Data Generation Approach

Abstract The sparsity of labeled data is an obstacle to the development of Relation Extraction (RE) models and the completion of databases in various biomedical areas. While being of high interest in drug-discovery, the literature on natural products, reporting the identification of potential bioactive compounds from organisms, is a concrete example of such an overlooked topic. To mark the start of this new task, we created the first curated evaluation dataset and extracted literature items from the LOTUS database to build training sets. To this end, we developed a new sampler, inspired by diversity metrics in ecology, named Greedy Maximum Entropy sampler (https://github.com/idiap/gme-sampler). The strategic optimization of both balance and diversity of the selected items in the evaluation set is important given the resource-intensive nature of manual curation. After quantifying the noise in the training set, in the form of discrepancies between the text of input abstracts and the expected output labels, we explored different strategies accordingly. Framing the task as an end-to-end Relation Extraction, we evaluated the performance of standard fine-tuning (BioGPT, GPT-2, and Seq2rel) and few-shot learning with open Large Language Models (LLMs) (LLaMA 7B-65B). In addition to their evaluation in few-shot settings, we explore the potential of open LLMs as synthetic data generators and propose a new workflow for this purpose. All evaluated models exhibited substantial improvements when fine-tuned on synthetic abstracts rather than the original noisy data.We provide our best performing (F1-score = 59:0) BioGPT-Large model for end-to-end RE of natural products relationships along with all the training and evaluation datasets. See more details at https://github.com/idiap/abroad-re.

A task‐centric knowledge graph construction method based on multi‐modal representation learning for industrial maintenance automation

AbstractMaintenance manuals are crucial information sources for maintenance and repair. Prior studies explored factual knowledge extraction from textual documents. However, maintenance knowledge in manuals is more task‐centric rather than factual knowledge and often documented in an unstructured Portable Document Format (PDF), posing challenges for knowledge extraction. Addressing this, this research develops effective methods to extract task‐centric maintenance knowledge from unstructured PDF manuals. A new Task‐centric Knowledge Graph (TCKG) schema centralized on maintenance task components (MTCs) is proposed to address the need for structured knowledge representation. A method (Heterogeneous Graph‐based Method, HGM) for knowledge extraction is then proposed, which is enhanced by incorporating visual and spatial information. In the experiments, the proposed HGM exhibits robust performance in the knowledge extraction process, surpassing the baseline Graph‐based Interaction Model with a Tracker (GIT) method in MTCs extraction by 13.3%, and the baseline Translate Embedding (TransE) method in MTCs' relation extraction by 3.8%. A series of ablation studies also prove that including visual and spatial information through the proposed method can improve the relation extraction performance by over 10%. This research supplies valuable insights for future developments in information extraction from maintenance manuals.

A large-scale dataset for korean document-level relation extraction from encyclopedia texts

Document-level relation extraction (RE) aims to predict the relational facts between two given entities from a document. Unlike widespread research on document-level RE in English, Korean document-level RE research is still at the very beginning due to the absence of a dataset. To accelerate the studies, we present TREK (Toward Document-Level Relation Extraction in Korean) dataset constructed from Korean encyclopedia documents written by the domain experts. We provide detailed statistical analyses for our large-scale dataset and human evaluation results suggest the assured quality of TREK . Also, we introduce the document-level RE model that considers the named entity-type while considering the Korean language’s properties. In the experiments, we demonstrate that our proposed model outperforms the baselines and conduct qualitative analysis.

Knowledge graph construction for heart failure using large language models with prompt engineering.

Constructing an accurate and comprehensive knowledge graph of specific diseases is critical for practical clinical disease diagnosis and treatment, reasoning and decision support, rehabilitation, and health management. For knowledge graph construction tasks (such as named entity recognition, relation extraction), classical BERT-based methods require a large amount of training data to ensure model performance. However, real-world medical annotation data, especially disease-specific annotation samples, are very limited. In addition, existing models do not perform well in recognizing out-of-distribution entities and relations that are not seen in the training phase. In this study, we present a novel and practical pipeline for constructing a heart failure knowledge graph using large language models and medical expert refinement. We apply prompt engineering to the three phases of schema design: schema design, information extraction, and knowledge completion. The best performance is achieved by designing task-specific prompt templates combined with the TwoStepChat approach. Experiments on two datasets show that the TwoStepChat method outperforms the Vanillia prompt and outperforms the fine-tuned BERT-based baselines. Moreover, our method saves 65% of the time compared to manual annotation and is better suited to extract the out-of-distribution information in the real world.

Interaction-and-Response Network for Distantly Supervised Relation Extraction.

Distantly supervised relation extraction (DSRE) aims to identify semantic relations from massive plain texts. A broad range of the prior research has leveraged a series of selective attention mechanisms over sentences in a bag to extract relation features without considering dependencies among the relation features. As a result, potential discriminative information existed in the dependencies is ignored, causing a decline in the performance of extracting entity relations. In this article, we focus on going beyond the selective attention mechanisms and propose a new framework termed interaction-and-response network (IR-Net) that adaptively recalibrates the features of sentence, bag, and group levels by explicitly modeling interdependencies among the features on each level. The IR-Net consists of a series of interactive and responsive modules throughout feature hierarchy, seeking to strengthen its power of learning salient discriminative features for distinguishing entity relations. We conduct extensive experiments on three benchmark DSRE datasets, including NYT-10, NYT-16, and Wiki-20m. The experimental results demonstrate that the IR-Net brings obvious improvements in performance when comparing ten state-of-the-art DSRE methods for entity relation extraction.