Entity Disambiguation Research Articles

Feature/vector entity retrieval and disambiguation techniques to create a supervised and unsupervised semantic table interpretation approach

Recently, there has been an increasing interest in extracting and annotating tables on the Web. This activity allows the transformation of textual data into machine-readable formats to enable the execution of various artificial intelligence tasks, e.g., semantic search and dataset extension. Semantic Table Interpretation (STI) is the process of annotating elements in a table. The paper explores Semantic Table Interpretation, addressing the challenges of Entity Retrieval and Entity Disambiguation in the context of Knowledge Graphs (KGs). It introduces LamAPI, an Information Retrieval system with string/type-based filtering and s-elBat, an Entity Disambiguation technique that combines heuristic and ML-based approaches. By applying the acquired know-how in the field and extracting algorithms, techniques and components from our previous STI approaches and the state of the art, we have created a new platform capable of annotating any tabular data, ensuring a high level of quality.

A graph based named entity disambiguation using clique partitioning and semantic relatedness

Disambiguating name mentions in texts is a crucial task in Natural Language Processing, especially in entity linking. The credibility and efficiency of such systems depend largely on this task. For a given name entity mention in a text, there are many potential candidate entities that may refer to it in the knowledge base. Therefore, it is very difficult to assign the correct candidate from the whole set of candidate entities of this mention. To solve this problem, collective entity disambiguation is a prominent approach. In this paper, we present a novel algorithm called CPSR for collective entity disambiguation, which is based on a graph approach and semantic relatedness. A clique partitioning algorithm is used to find the best clique that contains a set of candidate entities. These candidate entities provide the answers to the corresponding mentions in the disambiguation process. To evaluate our algorithm, we carried out a series of experiments on seven well-known datasets, namely, AIDA/CoNLL2003-TestB, IITB, MSNBC, AQUAINT, ACE2004, Cweb, and Wiki. The Kensho Derived Wikimedia Dataset (KDWD) is used as the knowledge base for our system. From the experimental results, our CPSR algorithm outperforms both the baselines and other well-known state-of-the-art approaches.

Entity linking for English and other languages: a survey

Extracting named entities text forms the basis for many crucial tasks such as information retrieval and extraction, machine translation, opinion mining, sentiment analysis and question answering. This paper presents a survey of the research literature on named entity linking, including named entity recognition and disambiguation. We present 200 works by focusing on 43 papers (5 surveys and 38 research works). We also describe and classify 56 resources, including 25 tools and 31 corpora. We focus on the most recent papers, where more than 95% of the described research works are after 2015. To show the efficiency of our construction methodology and the importance of this state of the art, we compare it to other surveys presented in the research literature, which were based on different criteria (such as the domain, novelty and presented models and resources). We also present a set of open issues (including the dominance of the English language in the proposed studies and the frequent use of NER rather than the end-to-end systems proposing NED and EL) related to entity linking based on the research questions that this survey aims to answer.

Enriching Portuguese Medieval Texts with Named Entity Recognition

Historical data poses unique challenges to natural language processing (NLP) and information retrieval (IR) tools, including digitization errors, lack of annotated data, and diachronic-specific issues. However, the increasing recognition of the value in historical documents has promoted efforts to semantically enrich and optimize their analysis. This article contributes to this endeavour by enriching the Corpus de Textos Antigos through NLP tools and techniques to enhance its usability and support research. The corpus undergoes linguistic annotation, including part-of-speech tagging, lemma annotation and named entity recognition (NER). Subsequently, the article delves into the tasks of entity disambiguation and entity linking, which involve identifying and disambiguating named entities by referring to a knowledge base (KB). Addressing the challenges posed by factors such as text state, epoch and the chosen KB, the article presents insights into related work, annotation results and the linguistic interest of a medieval annotated corpus for named entities. It concludes by discussing the challenges and providing avenues for future research in this domain.

Robust Chinese Short Text Entity Disambiguation Method Based on Feature Fusion and Contrastive Learning

To address the limitations of existing methods of short-text entity disambiguation, specifically in terms of their insufficient feature extraction and reliance on massive training samples, we propose an entity disambiguation model called COLBERT, which fuses LDA-based topic features and BERT-based semantic features, as well as using contrastive learning, to enhance the disambiguation process. Experiments on a publicly available Chinese short-text entity disambiguation dataset show that the proposed model achieves an F1-score of 84.0%, which outperforms the benchmark method by 0.6%. Moreover, our model achieves an F1-score of 74.5% with a limited number of training samples, which is 2.8% higher than the benchmark method. These results demonstrate that our model achieves better effectiveness and robustness and can reduce the burden of data annotation as well as training costs.

Leveraging Dual Gloss Encoders in Chinese Biomedical Entity Linking

Entity linking is the task of assigning a unique identity to named entities mentioned in a text, a sort of word sense disambiguation that focuses on automatically determining a pre-defined sense for a target entity to be disambiguated. This study proposes the DGE (Dual Gloss Encoders) model for Chinese entity linking in the biomedical domain. We separately model a dual encoder architecture, comprising a context-aware gloss encoder and a lexical gloss encoder, for contextualized embedding representations. DGE are then jointly optimized to assign the nearest gloss with the highest score for target entity disambiguation. The experimental datasets consist of a total of 10,218 sentences that were manually annotated with glosses defined in the BabelNet 5.0 across 40 distinct biomedical entities. Experimental results show that the DGE model achieved an F1-score of 97.81, outperforming other existing methods. A series of model analyses indicate that the proposed approach is effective for Chinese biomedical entity linking.

Knowledge Graph-Based Reinforcement Federated Learning for Chinese Question and Answering

Knowledge question and answering (Q&A) is widely used. However, most existing semantic parsing methods in Q&A usually use cascading, which can incur error accumulation. In addition, using only one institution’s Q&A data definitely will limit the Q&A performance, while data privacy prevents sharing between institutions. This article proposes a knowledge graph-based reinforcement federated learning (KGRFL)-based Q&A approach to address these challenges. We design an end-to-end multitask semantic parsing model MSP-bidirectional and auto-regressive transformers (BART) that identifies question categories while converting questions into SPARQL statements to improve semantic parsing. Meanwhile, a reinforcement learning (RL)-based model fusion strategy is proposed to improve the effectiveness of federated learning, which enables multi-institution joint modeling and data privacy protection using cross-domain knowledge. In particular, it also reduces the negative impact of low-quality clients on the global model. Furthermore, a prompt learning-based entity disambiguation method is proposed to address the semantic ambiguity problem because of joint modeling. The experiments show that the proposed method performs well on different datasets. The Q&A results of the proposed approach outperform the approach of using only a single institution. Experiments also demonstrate that the proposed approach is resilient to security attacks, which is required for real applications.

Improving entity linking by combining semantic entity embeddings and cross-attention encoder

Entity linking is an important task for information retrieval and knowledge graph construction. Most existing methods use a bi-encoder structure to encode mentions and entities in the same space, and learn contextual features for entity linking. However, this type of system still faces some problems: (1) the entity embedding part of the model only learns from the local context of the target entity, which is too unique for entity linking model to learn the context commonality of information; (2) the entity disambiguation part only uses similarity calculation once to determine the target entity, resulting in insufficient interaction between the mentions and candidate entities, and ineffective recall of real entities. We propose a new entity linking model based on graph neural network. Different from other bi-encoder retrieval systems, this paper introduces a fine-grained semantic enhancement information into the entity embedding part of the bi-encoder to reduce the specificity of the model. Then, the cross-attention encoder is used to re-rank the target mention and each candidate entity after the entity retrieval model. Experimental results show that although the model is not optimal in inference speed, it outperforms all baseline methods on the AIDA-CoNLL dataset, and has good generalization effects on four datasets in different fields such as MSNBC and ACE2004.

Self-Supervised Enhancement for Named Entity Disambiguation via Multimodal Graph Convolution.

Named entity disambiguation (NED) finds the specific meaning of an entity mention in a particular context and links it to a target entity. With the emergence of multimedia, the modalities of content on the Internet have become more diverse, which poses difficulties for traditional NED, and the vast amounts of information make it impossible to manually label every kind of ambiguous data to train a practical NED model. In response to this situation, we present MMGraph, which uses multimodal graph convolution to aggregate visual and contextual language information for accurate entity disambiguation for short texts, and a self-supervised simple triplet network (SimTri) that can learn useful representations in multimodal unlabeled data to enhance the effectiveness of NED models. We evaluated these approaches on a new dataset, MMFi, which contains multimodal supervised data and large amounts of unlabeled data. Our experiments confirm the state-of-the-art performance of MMGraph on two widely used benchmarks and MMFi. SimTri further improves the performance of NED methods. The dataset and code are available at https://github.com/LanceZPF/NNED_MMGraph.

Abstract 17760: Conversion of Dicom ECG Images to Tabular Format for Building Large Language Model in Diagnoses and Disease Progression of Cardiovascular Conditions

Introduction: DICOM Electrocardiography (ECG) images from individuals are routinely stored at the institutional PACS server, including normal and abnormal findings. Hypothesis: The study’s objective is to create a database of ECGs as per Standard Communication Protocol (SCP) and to build an accurate Large Language Model (LLM) across different cardiovascular conditions Methods: DICOM ECGs are retrieved, anonymized and labelled to convert the signals to (x, y) coordinates with the help of PyDicom libraries. The tabular format is then stored with lead definition and dimensions of duration and amplitude, metadata, clinical conditions, and textual diagnoses. Results: In the pilot study, 1308 ECGs from individuals with different age (<45 & >45 years), gender (male & female), binary clinical categories (normal & abnormal) and heart rate (<90 & >90 per min) are selected for analysis and modeling. The DICOM image signal reproducibility following conversion (Figure 1 & 2) shows Cross Correlation Percentage of 0.85 (0.76 - 0.94) (Figure 3). An eXtreme Gradient Boosting (XGBoost) model (6) was trained to predict above variables. The results (Figure 4) show AUC highest for predicting binary clinical categories (normal vs abnormal) at 0.81 (0.7 - 0.9) and lowest for predicting age (<45 & >45 years) at 0.6 (0.46 - 0.7). The accuracies in the leads also show similar trend. The limitations include reproducibility of results in Lead VIII, IX and XI, which are sub average and hence being retrained with optimizers. These are initial results with relatively smaller database where multiclass disease or condition classifications are not performed. Conclusions: To conclude, the methodology possesses opportunities to improve the model with Deep Learning and Entity Disambiguation Techniques (NLP) to build Large Language ECG Models for cardiovascular diagnoses and disease progression trajectories.

Adaptive deep learning for entity disambiguation via knowledge-based risk analysis

The state-of-the-art performance on entity disambiguation has been reached by deep neural networks. However, the task remains very challenging due to the complexity of natural language. Moreover, the target data distribution is often different from that of training data. In this paper, we address the limitation of deep entity disambiguation from the perspective of misprediction risk. We propose a knowledge-based approach of risk analysis for entity disambiguation, and leverage it to enable adaptive deep learning. The proposed approach generates risk features by extracting evidences from the knowledge base, and then models them as a linearly-weighted random vector where an attention mechanism is used to focus on the most significant components. Finally, it estimates misprediction risk of the aggregated probability distribution via the Conditional Value-at-Risk metric. Furthermore, we demonstrate how to utilize risk analysis results in adaptive deep learning via two-phase training, the first phase fits on labeled training data while the second one minimizes misprediction risk on unlabeled target data. We evaluate the performance of the proposed approach on benchmark datasets through a comparative study. Our thorough experiments demonstrate that it can detect mispredictions more accurately than existing alternatives and can substantially improve the performance of deep learning models.

Unifying context with labeled property graph: A pipeline-based system for comprehensive text representation in NLP

Extracting valuable insights from vast amounts of unstructured digital text presents significant challenges across diverse domains. This research addresses this challenge by proposing a novel pipeline-based system that generates domain-agnostic and task-agnostic text representations. The proposed approach leverages labeled property graphs (LPG) to encode contextual information, facilitating the integration of diverse linguistic elements into a unified representation. The proposed system enables efficient graph-based querying and manipulation by addressing the crucial aspect of comprehensive context modeling and fine-grained semantics. The effectiveness of the proposed system is demonstrated through the implementation of NLP components that operate on LPG-based representations. Additionally, the proposed approach introduces specialized patterns and algorithms to enhance specific NLP tasks, including nominal mention detection, named entity disambiguation, event enrichments, event participant detection, and temporal link detection. The evaluation of the proposed approach, using the MEANTIME corpus comprising manually annotated documents, provides encouraging results and valuable insights into the system's strengths. The proposed pipeline-based framework serves as a solid foundation for future research, aiming to refine and optimize LPG-based graph structures to generate comprehensive and semantically rich text representations, addressing the challenges associated with efficient information extraction and analysis in NLP.

Disambiguation of company names via deep recurrent networks

Name Entity Disambiguation is the Natural Language Processing task of identifying textual records corresponding to the same Named Entity, i.e., real-world entities represented as a list of attributes (names, places, organisations, etc.). In this work, we face the task of disambiguating companies on the basis of their written names. We propose a Siamese LSTM Network approach to extract – via supervised learning – an embedding of company name strings in a (relatively) low dimensional vector space and use this representation to identify pairs of company names that actually represent the same company (i.e., the same Entity).Given that the manual labelling of string pairs is a rather onerous task, we analyse how an Active Learning approach to prioritise the samples to be labelled leads to a more efficient overall learning pipeline.The contributions of this work are: with empirical investigations on real-world industrial data, we show that our proposed Siamese Network outperforms several benchmark approaches based on standard string matching algorithms when enough labelled data are available; moreover, we show that Active Learning prioritisation is indeed helpful when labelling resources are limited, and let the learning models reach the out-of-sample performance saturation with less labelled data with respect to standard (random) data labelling approaches.

DomainNet: Homograph Detection and Understanding in Data Lake Disambiguation

Modern data lakes are heterogeneous in the vocabulary that is used to describe data. We study a problem of disambiguation in data lakes: How can we determine if a data value occurring more than once in the lake has different meanings and is therefore a homograph? While word and entity disambiguation have been well studied in computational linguistics, data management, and data science, we show that data lakes provide a new opportunity for disambiguation of data values, because tables implicitly define a massive network of interconnected values. We introduce DomainNet , which efficiently represents this network, and investigate to what extent it can be used to disambiguate values without requiring any supervision. DomainNet leverages network-centrality measures on a bipartite graph whose nodes represent data values and attributes to determine if a value is a homograph. A thorough experimental evaluation demonstrates that state-of-the-art techniques in domain discovery cannot be re-purposed to compete with our method. Specifically, using a domain discovery method to identify homographs achieves an F1-score of 0.38 versus 0.69 for DomainNet , which separates homographs well from data values that have a unique meaning. On a real data lake, our top-100 precision is 93%. Given a homograph, we also present a novel method for determining the number of meanings of the homograph and for assigning its data lake attributes to a meaning. We show the influence of homographs on two downstream tasks: entity-matching and domain discovery.

Entity generation algorithm based on reference expansion

The extraction and understanding of text knowledge become increasingly crucial in the age of big data. One of the current research areas in the field of natural language processing (NLP) is how to accurately understand the text and collect accurate linguistic information because Chinese vocabulary is diverse and ambiguous. This paper mainly studies the candidate entity generation module of the entity link system. The candidate entity generation module constructs an entity reference expansion algorithm to improve the recall rate of candidate entities. In order to improve the efficiency of the connection algorithm of the entire system while ensuring the recall rate of candidate entities, we design a graph model filtering algorithm that fuses shallow semantic information to filter the list of candidate entities, and verify and analyze the efficiency of the algorithm through experiments. By analyzing the related technology of the entity linking algorithm, we study the related technology of candidate entity generation and entity disambiguation, improve the traditional entity linking algorithm, and give an innovative and practical entity linking model. The recall rate exceeds 82%, and the link accuracy rate exceeds 73%. Efficient and accurate entity linking can help machines to better understand text semantics, further promoting the development of NLP and improving the users’ knowledge acquisition experience on the text.

A semantic union model for open domain Chinese knowledge base question answering

In Open-domain Chinese Knowledge Base Question Answering (ODCKBQA), most common simple questions can be answered by a single relational fact in the knowledge base (KB). The abbreviations, aliases, and nesting of entities in Chinese question sentences, and the gap between them and the structured semantics in the knowledge base, make it difficult for the system to accurately return answers. This study proposes a semantic union model (SUM), which concatenates candidate entities and candidate relationships, using a contrastive learning algorithm to learn the semantic vector representation of question and candidate entity-relation pairs, and perform cosine similarity calculations to simultaneously complete entity disambiguation and relation matching tasks. It can provide information for entity disambiguation through the relationships between entities, avoid error propagation, and improve the system performance. The experimental results show that the system achieves a good average F1 of 85.94% on the dataset provided by the NLPCC-ICCPOL 2016 KBQA task.

End-to-End Entity Linking with Hierarchical Reinforcement Learning

Entity linking (EL) is the task of linking the text segments to the referring entities in the knowledge graph, typically decomposed into mention detection, and entity disambiguation. Compared to traditional methods treating the two tasks separately, recent end-to-end entity linking methods exploit the mutual dependency between mentions and entities to achieve better performance. However, existing end-to-end EL methods have problems utilizing the dependency of mentions and entities in the task. To this end, we propose to model the EL task as a hierarchical decision-making process and design a hierarchical reinforcement learning algorithm to solve the problem. We conduct extensive experiments to show that the proposed method achieves state-of-the-art performance in several EL benchmark datasets. Our code is publicly available at https://github.com/lhlclhl/he2eel.

Semantic Tagging for the Urdu Language: Annotated Corpus and Multi-Target Classification Methods

Extracting and analysing meaning-related information from natural language data has attracted the attention of researchers in various fields, such as natural language processing, corpus linguistics, information retrieval, and data science. An important aspect of such automatic information extraction and analysis is the annotation of language data using semantic tagging tools. Different semantic tagging tools have been designed to carry out various levels of semantic analysis, for instance, named entity recognition and disambiguation, sentiment analysis, word sense disambiguation, content analysis, and semantic role labelling. Common to all of these tasks, in the supervised setting, is the requirement for a manually semantically annotated corpus, which acts as a knowledge base from which to train and test potential word and phrase-level sense annotations. Many benchmark corpora have been developed for various semantic tagging tasks, but most are for English and other European languages. There is a dearth of semantically annotated corpora for the Urdu language, which is widely spoken and used around the world. To fill this gap, this study presents a large benchmark corpus and methods for the semantic tagging task for the Urdu language. The proposed corpus contains 8,000 tokens in the following domains or genres: news, social media, Wikipedia, and historical text (each domain having 2K tokens). The corpus has been manually annotated with 21 major semantic fields and 232 sub-fields with the USAS (UCREL Semantic Analysis System) semantic taxonomy which provides a comprehensive set of semantic fields for coarse-grained annotation. Each word in our proposed corpus has been annotated with at least one and up to nine semantic field tags to provide a detailed semantic analysis of the language data, which allowed us to treat the problem of semantic tagging as a supervised multi-target classification task. To demonstrate how our proposed corpus can be used for the development and evaluation of Urdu semantic tagging methods, we extracted local, topical and semantic features from the proposed corpus and applied seven different supervised multi-target classifiers to them. Results show an accuracy of 94% on our proposed corpus which is free and publicly available to download.

Entity Linking Method for Chinese Short Texts with Multiple Embedded Representations

Entity linking, a crucial task in the realm of natural language processing, aims to link entity mentions in a text to their corresponding entities in the knowledge base. While long documents provide abundant contextual information, facilitating feature extraction for entity identification and disambiguation, entity linking in Chinese short texts presents significant challenges. This study introduces an innovative approach to entity linking within Chinese short texts, combining multiple embedding representations. It integrates embedding representations from both entities and relations in the knowledge graph triples, as well as embedding representations from the descriptive text of entities and relations, to enhance the performance of entity linking. The method also incorporates external semantic supplements to strengthen the model’s feature learning capabilities. The Multi-Embedding Representation–Bidirectional Encoder Representation from Transformers–Bidirectional Gated Recurrent Unit (MER-BERT-BiGRU) neural network model is employed for embedding learning. The precision, recall, and F1 scores reached 89.73%, 92.18%, and 90.94% respectively, demonstrating the effectiveness of our approach.

A novel text representation which enables image classifiers to also simultaneously classify text, applied to name disambiguation

We introduce a novel method for converting text data into abstract image representations, which allows image-based processing techniques (e.g. image classification networks) to be applied to text-based comparison problems. We apply the technique to entity disambiguation of inventor names in US patents, obtaining a list of IDs which identify individual inventors with high accuracy. The method involves converting text from each pairwise comparison between two inventor name records into a 2D RGB (stacked) image representation. We then train an image classification neural network to discriminate between such pairwise comparison images. The trained neural network then labels each pair of records as either matched (same inventor) or non-matched (different inventors), producing highly accurate results. Our new text-to-image representation method could also be used more broadly for other text comparison problems, such as entity disambiguation of academic publications, or for problems that require simultaneous classification of both text and image datasets.