Bidirectional Long Short-term Memory Conditional Random Field Research Articles

A knowledge graph and BiLSTM-CRF-enabled intelligent adaptive learning model and its potential application

The significant advancement of Artificial Intelligence (AI) technology has led to the gradual emergence of intelligent adaptive learning as a pivotal modality for acquiring knowledge. The knowledge graph, as a crucial component of the intelligent adaptive learning model, assumes the vital responsibility of guiding and facilitating the learning process while evaluating its outcomes. The current knowledge graph, however, presents several challenges including fragmented knowledge content, inadequate ability characterization, and limited automatic construction capabilities, which hinder its effectiveness in assisting learners to achieve their learning objectives. The paper proposes the knowledge-resource-objective (KRO) model of knowledge graph, based on a robust correlation between knowledge, learning resources, and learning objectives. In the KRO model, Bidirectional Encoder Representations from Transformers (BERT) is utilized for data pre-processing. Protége is employed for ontology construction. Entities and relations are synchronously annotated using the Main-Entity+ Relation +Begin-Inside-End-Single-Other (ME+R+BIESO) corpus labeling model to complete the triplet construction. Additionally, entities and relations are extracted synchronously utilizing a Bi-directional Long-Short Term Memory-Conditional Random Field (BiLSTM-CRF) model. Knowledge visualization and reasoning are achieved through Neo4j graph database. The study employs the Letter of Credit as a case study to examine the application of the KRO model, and assesses its efficacy through metrics such as accuracy, precision, recall, and F1-score. The research demonstrates that the KRO model effectively targets learning objectives, restructures knowledge and resources, optimizes learning pathways, provides recommendations for learning resources, and visualizes learners' cognitive states. As a result, it enhances the effectiveness of intelligent adaptive learning. The subsequent research will continue to focus on the visual representation of learners' learning progress and explore the practical application of learners' profiles in diverse disciplines, aiming to facilitate the realization of personalized learning for individuals.

Web-Based Application Based on Human-in-the-Loop Deep Learning for Deidentifying Free-Text Data in Electronic Medical Records: Development and Usability Study.

The narrative free-text data in electronic medical records (EMRs) contain valuable clinical information for analysis and research to inform better patient care. However, the release of free text for secondary use is hindered by concerns surrounding personally identifiable information (PII), as protecting individuals' privacy is paramount. Therefore, it is necessary to deidentify free text to remove PII. Manual deidentification is a time-consuming and labor-intensive process. Numerous automated deidentification approaches and systems have been attempted to overcome this challenge over the past decade. We sought to develop an accurate, web-based system deidentifying free text (DEFT), which can be readily and easily adopted in real-world settings for deidentification of free text in EMRs. The system has several key features including a simple and task-focused web user interface, customized PII types, use of a state-of-the-art deep learning model for tagging PII from free text, preannotation by an interactive learning loop, rapid manual annotation with autosave, support for project management and team collaboration, user access control, and central data storage. DEFT comprises frontend and backend modules and communicates with central data storage through a filesystem path access. The frontend web user interface provides end users with a user-friendly workspace for managing and annotating free text. The backend module processes the requests from the frontend and performs relevant persistence operations. DEFT manages the deidentification workflow as a project, which can contain one or more data sets. Customized PII types and user access control can also be configured. The deep learning model is based on a Bidirectional Long Short-Term Memory-Conditional Random Field (BiLSTM-CRF) with RoBERTa as the word embedding layer. The interactive learning loop is further integrated into DEFT to speed up the deidentification process and increase its performance over time. DEFT has many advantages over existing deidentification systems in terms of its support for project management, user access control, data management, and an interactive learning process. Experimental results from DEFT on the 2014 i2b2 data set obtained the highest performance compared to 5 benchmark models in terms of microaverage strict entity-level recall and F1-scores of 0.9563 and 0.9627, respectively. In a real-world use case of deidentifying clinical notes, extracted from 1 referral hospital in Sydney, New South Wales, Australia, DEFT achieved a high microaverage strict entity-level F1-score of 0.9507 on a corpus of 600 annotated clinical notes. Moreover, the manual annotation process with preannotation demonstrated a 43% increase in work efficiency compared to the process without preannotation. DEFT is designed for health domain researchers and data custodians to easily deidentify free text in EMRs. DEFT supports an interactive learning loop and end users with minimal technical knowledge can perform the deidentification work with only a shallow learning curve.

A cross-institutional evaluation on breast cancer phenotyping NLP algorithms on electronic health records

ObjectiveTransformer-based language models are prevailing in the clinical domain due to their excellent performance on clinical NLP tasks. The generalizability of those models is usually ignored during the model development process. This study evaluated the generalizability of CancerBERT, a Transformer-based clinical NLP model, along with classic machine learning models, i.e., conditional random field (CRF), bi-directional long short-term memory CRF (BiLSTM-CRF), across different clinical institutes through a breast cancer phenotype extraction task. Materials and methodsTwo clinical corpora of breast cancer patients were collected from the electronic health records from the University of Minnesota (UMN) and Mayo Clinic (MC), and annotated following the same guideline. We developed three types of NLP models (i.e., CRF, BiLSTM-CRF and CancerBERT) to extract cancer phenotypes from clinical texts. We evaluated the generalizability of models on different test sets with different learning strategies (model transfer vs locally trained). The entity coverage score was assessed with their association with the model performances. ResultsWe manually annotated 200 and 161 clinical documents at UMN and MC. The corpora of the two institutes were found to have higher similarity between the target entities than the overall corpora. The CancerBERT models obtained the best performances among the independent test sets from two clinical institutes and the permutation test set. The CancerBERT model developed in one institute and further fine-tuned in another institute achieved reasonable performance compared to the model developed on local data (micro-F1: 0.925 vs 0.932). ConclusionsThe results indicate the CancerBERT model has superior learning ability and generalizability among the three types of clinical NLP models for our named entity recognition task. It has the advantage to recognize complex entities, e.g., entities with different labels.

Enhancement of Question Answering System Accuracy via Transfer Learning and BERT

Entity linking and predicate matching are two core tasks in the Chinese Knowledge Base Question Answering (CKBQA). Compared with the English entity linking task, the Chinese entity linking is extremely complicated, making accurate Chinese entity linking difficult. Meanwhile, strengthening the correlation between entities and predicates is the key to the accuracy of the question answering system. Therefore, we put forward a Bidirectional Encoder Representation from Transformers and transfer learning Knowledge Base Question Answering (BAT-KBQA) framework, which is on the basis of feature-enhanced Bidirectional Encoder Representation from Transformers (BERT), and then perform a Named Entity Recognition (NER) task, which is appropriate for Chinese datasets using transfer learning and the Bidirectional Long Short-Term Memory-Conditional Random Field (BiLSTM-CRF) model. We utilize a BERT-CNN (Convolutional Neural Network) model for entity disambiguation of the problem and candidate entities; based on the set of entities and predicates, a BERT-Softmax model with answer entity predicate features is introduced for predicate matching. The answer ultimately chooses to integrate entities and predicates scores to determine the definitive answer. The experimental results indicate that the model, which is developed by us, considerably enhances the overall performance of the Knowledge Base Question Answering (KBQA) and it has the potential to be generalizable. The model also has better performance on the dataset supplied by the NLPCC-ICCPOL2016 KBQA task with a mean F1 score of 87.74% compared to BB-KBQA.

Military Chain: Construction of Domain Knowledge Graph of Kill Chain Based on Natural Language Model

With the advent of the Big Data era, the specialized data in the kill chain domain has increased dramatically, and the engine-based method of retrieving information can hardly meet the users' need for more accurate answers. The kill chain domain includes four components: control equipment, sensor equipment, strike equipment (weapon and platform), and evaluator equipment, as well as related data which contain a large amount of valuable information such as the parameter information contained in each component. If these fragmented and confusing data are integrated and effective query methods are established, they can help professionals complete the military kill chain knowledge system. The knowledge system constructed in this paper is based on the Neo4j graph database and the US Command simulation system to establish a target-oriented knowledge map of kill chain, aiming to provide data support for the Q&A system. Secondly, in order to facilitate the query, this paper establishes entity and relationship/attribute mining based on the continuous bag-of-words (CBOW) encoding model, bidirectional long short-term memory–conditional random field (BiLSTM-CRF) named entity model, and bidirectional gated recurrent neural network (BiGRU) intent recognition model for Chinese kill chain question and answer; returns the corresponding entity or attribute values in combination with the knowledge graph triad form; and finally constructs the answer return. The constructed knowledge map of the kill chain contains 2767 items (including sea, land, and air), and the number of parameters involved is 30124. The number of model parameters of the deep learning network is 27.9 M for the Q&A system built this time, and the accuracy rate is 85.5% after 200 simulated queries.

Leveraging Integrated Learning for Open-Domain Chinese Named Entity Recognition

Named entity recognition (NER) is a fundamental technique in natural language processing that provides preconditions for tasks, such as natural language question reasoning, text matching, and semantic text similarity. Compared to English, the challenge of Chinese NER lies in the noise impact caused by the complex meanings, diverse structures, and ambiguous semantic boundaries of the Chinese language itself. At the same time, compared with specific domains, open-domain entity types are more complex and changeable, and the number of entities is considerably larger. Thus, the task of Chinese NER is more difficult. However, existing open-domain NER methods have low recognition rates. Therefore, this paper proposes a method based on the bidirectional long short-term memory conditional random field (BiLSTM-CRF) model, which leverages integrated learning to improve the efficiency of Chinese NER. Compared with single models, including CRF, BiLSTM-CRF, and gated recurrent unit-CRF, the proposed method can significantly improve the accuracy of open-domain Chinese NER.

Chinese Named Entity Recognition Based on Knowledge Based Question Answering System

The KBQA (Knowledge-Based Question Answering) system is an essential part of the smart customer service system. KBQA is a type of QA (Question Answering) system based on KB (Knowledge Base). It aims to automatically answer natural language questions by retrieving structured data stored in the knowledge base. Generally, when a KBQA system receives the user’s query, it first needs to recognize topic entities of the query, such as name, location, organization, etc. This process is the NER (Named Entity Recognition). In this paper, we use the Bidirectional Long Short-Term Memory-Conditional Random Field (Bi-LSTM-CRF) model and introduce the SoftLexicon method for a Chinese NER task. At the same time, according to the analysis of the characteristics of application scenario, we propose a fuzzy matching module based on the combination of multiple methods. This module can efficiently modify the error recognition results, which can further improve the performance of entity recognition. We combine the NER model and the fuzzy matching module into an NER system. To explore the availability of the system in some specific fields, such as a power grid field, we utilize the power grid-related original data collected by the Hebei Electric Power Company to improve our system according to the characteristics of data in the power grid field. We innovatively make the dataset and high-frequency word lexicon in the power grid field, which makes our proposed NER system perform better in recognizing entities in the field of power grid. We used the cross-validation method for validation. The experimental results show that the F1-score of the improved NER model on the power grid dataset reaches 92.43%. After processing the recognition results by using the fuzzy matching module, about 99% of the entities in the test set can be correctly recognized. It proves that the proposed NER system can achieve excellent performance in the application scenario of a power grid. The results of this work will also fill the gap in the research of intelligent customer-service-related technologies in the power grid field in China.

MEduKG: A Deep-Learning-Based Approach for Multi-Modal Educational Knowledge Graph Construction

The popularity of information technology has given rise to a growing interest in smart education and has provided the possibility of combining online and offline education. Knowledge graphs, an effective technology for knowledge representation and management, have been successfully utilized to manage massive educational resources. However, the existing research on constructing educational knowledge graphs ignores multiple modalities and their relationships, such as teacher speeches and their relationship with knowledge. To tackle this problem, we propose an automatic approach to construct multi-modal educational knowledge graphs that integrate speech as a modal resource to facilitate the reuse of educational resources. Specifically, we first propose a fine-tuned Bidirectional Encoder Representation from Transformers (BERT) model based on education lexicon, called EduBERT, which can adaptively capture effective information in the education field. We also add a Bidirectional Long Short-Term Memory-Conditional Random Field (BiLSTM-CRF) to effectively identify educational entities. Then, the locational information of the entity is incorporated into BERT to extract the educational relationship. In addition, to cover the shortage of traditional text-based knowledge graphs, we focus on collecting teacher speech to construct a multi-modal knowledge graph. We propose a speech-fusion method that links these data into the graph as a class of entities. The numeric results show that our proposed approach can manage and present various modes of educational resources and that it can provide better education services.

Predicting Drug Indications and Side Effects Using Deep Learning and Transfer Learning

In the area of biology, text mining is commonly used since it obtains the unknown relationship among medicines, phenotypes and syndromes from much information. Enhanced Topic modeling with Improved Predict drug Indications and Side effects using Topic modelling and Natural language processing (ETP-IPISTON) has been employed to predict the drug-phenotype and drug-side effect association. Initially, corpus documents are collected from the literature data and the topics in the data are modeled using logistic Linear Discriminative Analysis (LDA) and Bi-directional Long-Short Term Memory-Conditional Random Field (BILSTM-CRF). From the sentences in the literature data, a dependency graph was constructed which discovered the relations between gene and drug. The product of the drug on phenotype rule was identified by the Gene Regulation Score (GRS) which creates the drug-topic probability matrix. The probability matrix and a syntactic distance measure was processed in Classification and Regression Tree (CART), Naïve Bayes (NB), logistic regression and Convolutional Neural Network (CNN) classifiers for estimating the drug-gene and drug-side effects. Besides the literature data, social media offers various promising resources with massive volume of data that can be useful in the drug-phenotype and drug-side effect association prediction. So in this paper, drug information with gene, disease and side effects are extracted from different social media such as Twitter, Facebook and LinkedIn and it can be used with the literature data to provide more relevant disease and drug relations. In addition to this, topic modeling with transfer learning is introduced to consider the element categories, probability of overlapping elements and deep contextual significance of a text for better modeling of topics. The topic modeling with transfer learning shares as much knowledge as possible between the literature data and social media information for topic modeling. The topics from social media and literature data are used for creating the drug-topic matrix. The probability matrix and syntactic distance measure are given as input to CART, NB, logistic regression and CNN for estimating the drug-gene and drug-side effect association. This proposed work is named as Enhanced Topic Modeling with Transfer Leaning- IPISTON (ETPTL-IPISTON). The simulation findings exhibit that the efficiency of ETPTL-IPISTON than the traditional methods.

Use of BERT (Bidirectional Encoder Representations from Transformers)-Based Deep Learning Method for Extracting Evidences in Chinese Radiology Reports: Development of a Computer-Aided Liver Cancer Diagnosis Framework.

BackgroundLiver cancer is a substantial disease burden in China. As one of the primary diagnostic tools for detecting liver cancer, dynamic contrast-enhanced computed tomography provides detailed evidences for diagnosis that are recorded in free-text radiology reports.ObjectiveThe aim of our study was to apply a deep learning model and rule-based natural language processing (NLP) method to identify evidences for liver cancer diagnosis automatically.MethodsWe proposed a pretrained, fine-tuned BERT (Bidirectional Encoder Representations from Transformers)-based BiLSTM-CRF (Bidirectional Long Short-Term Memory-Conditional Random Field) model to recognize the phrases of APHE (hyperintense enhancement in the arterial phase) and PDPH (hypointense in the portal and delayed phases). To identify more essential diagnostic evidences, we used the traditional rule-based NLP methods for the extraction of radiological features. APHE, PDPH, and other extracted radiological features were used to design a computer-aided liver cancer diagnosis framework by random forest.ResultsThe BERT-BiLSTM-CRF predicted the phrases of APHE and PDPH with an F1 score of 98.40% and 90.67%, respectively. The prediction model using combined features had a higher performance (F1 score, 88.55%) than those using APHE and PDPH (84.88%) or other extracted radiological features (83.52%). APHE and PDPH were the top 2 essential features for liver cancer diagnosis.ConclusionsThis work was a comprehensive NLP study, wherein we identified evidences for the diagnosis of liver cancer from Chinese radiology reports, considering both clinical knowledge and radiology findings. The BERT-based deep learning method for the extraction of diagnostic evidence achieved state-of-the-art performance. The high performance proves the feasibility of the BERT-BiLSTM-CRF model in information extraction from Chinese radiology reports. The findings of our study suggest that the deep learning–based method for automatically identifying evidences for diagnosis can be extended to other types of Chinese clinical texts.

Named Entity Recognition in Equipment Support Field Using Tri-Training Algorithm and Text Information Extraction Technology

Weaponry equipment names belong to an important military naming entity that is difficult to identify because of features, such as complex components, miscellaneous, and scarce annotation corpus. Here, the automatic recognition of weaponry equipment names is specifically explored, a NER (Named Entity Recognition) algorithm is proposed based on BI-LSTM-CRF (Bi-directional Long Short Term Memory Conditional Random Field), thereby demonstrating the effectiveness of domain features in domain-specific entity recognition. Firstly, Chinese characters are represented by word embedding and input into the model. Then, the input feature vector sequence is processed by BI-LSTM (Bi-directional Long Short Term Memory) NN (Neural Network) to extract context semantic learning features. Finally, the learned features are connected to the linear CRF (Conditional Random Field), the NEs (Named Entities) in the equipment support field are labeled, and the NER results are obtained and output. The experimental results show that the accuracy of the NER algorithm based on the BI-LSTM-CRF model is 92.02%, the recall rate is 93.21%, and the F1 value reaches 93.88%. The effect of this model is better than the BI-LSTM NN model and LSTM-CRF (Long Short Term Memory Conditional Random Field) NN model. The proposed model provides some references for entity recognition in the field of equipment support.

Deep learning approaches for extracting adverse events and indications of dietary supplements from clinical text.

We sought to demonstrate the feasibility of utilizing deep learning models to extract safety signals related to the use of dietary supplements (DSs) in clinical text. Two tasks were performed in this study. For the named entity recognition (NER) task, Bi-LSTM-CRF (bidirectional long short-term memory conditional random field) and BERT (bidirectional encoder representations from transformers) models were trained and compared with CRF model as a baseline to recognize the named entities of DSs and events from clinical notes. In the relation extraction (RE) task, 2 deep learning models, including attention-based Bi-LSTM and convolutional neural network as well as a random forest model were trained to extract the relations between DSs and events, which were categorized into 3 classes: positive (ie, indication), negative (ie, adverse events), and not related. The best performed NER and RE models were further applied on clinical notes mentioning 88 DSs for discovering DSs adverse events and indications, which were compared with a DS knowledge base. For the NER task, deep learning models achieved a better performance than CRF, with F1 scores above 0.860. The attention-based Bi-LSTM model performed the best in the RE task, with an F1 score of 0.893. When comparing DS event pairs generated by the deep learning models with the knowledge base for DSs and event, we found both known and unknown pairs. Deep learning models can detect adverse events and indication of DSs in clinical notes, which hold great potential for monitoring the safety of DS use.

Delayed Combination of Feature Embedding in Bidirectional LSTM CRF for NER

Named Entity Recognition (NER) plays a vital role in natural language processing (NLP). Currently, deep neural network models have achieved significant success in NER. Recent advances in NER systems have introduced various feature selections to identify appropriate representations and handle Out-Of-the-Vocabulary (OOV) words. After selecting the features, they are all concatenated at the embedding layer before being fed into a model to label the input sequences. However, when concatenating the features, information collisions may occur and this would cause the limitation or degradation of the performance. To overcome the information collisions, some works tried to directly connect some features to latter layers, which we call the delayed combination and show its effectiveness by comparing it to the early combination. As feature encodings for input, we selected the character-level Convolutional Neural Network (CNN) or Long Short-Term Memory (LSTM) word encoding, the pre-trained word embedding, and the contextual word embedding and additionally designed CNN-based sentence encoding using a dictionary. These feature encodings are combined at early or delayed position of the bidirectional LSTM Conditional Random Field (CRF) model according to each feature’s characteristics. We evaluated the performance of this model on the CoNLL 2003 and OntoNotes 5.0 datasets using the F1 score and compared the delayed combination model with our own implementation of the early combination as well as the previous works. This comparison convinces us that our delayed combination is more effective than the early one and also highly competitive.

Earthquake Knowledge Graph Constructing Based on Social Intercourse Using BiLSTM-CRF

Nowadays, the earthquake has become a very serious topic. The earthquake-related information always appears first in social media. Constructing an earthquake knowledge graph can help dealing with earthquake news social media text data. This paper proposed the BiLSTM-CRF (Bi-directional Long Short Term Memory-Conditional Random Field) model to construct earthquake knowledge graph using news text data. The BiLSTM-CRF model identified the entities and then writes entities and their types to the table so that the entities and relationships between entities can be extracted for earthquake. The entities and relationships between entities have been combined with the form of RDF (Resource Description Framework) to construct earthquake knowledge graph on Neo4j database.

Feature Extraction and Analysis of Natural Language Processing for Deep Learning English Language

NLP (Natural Language Processing) is a technology that enables computers to understand human languages. Deep-level grammatical and semantic analysis usually uses words as the basic unit, and word segmentation is usually the primary task of NLP. In order to solve the practical problem of huge structural differences between different data modalities in a multi-modal environment and traditional machine learning methods cannot be directly applied, this paper introduces the feature extraction method of deep learning and applies the ideas of deep learning to multi-modal feature extraction. This paper proposes a multi-modal neural network. For each mode, there is a multilayer sub-neural network with an independent structure corresponding to it. It is used to convert the features in different modes to the same-modal features. In terms of word segmentation processing, in view of the problems that existing word segmentation methods can hardly guarantee long-term dependency of text semantics and long training prediction time, a hybrid network English word segmentation processing method is proposed. This method applies BI-GRU (Bidirectional Gated Recurrent Unit) to English word segmentation, and uses the CRF (Conditional Random Field) model to annotate sentences in sequence, effectively solving the long-distance dependency of text semantics, shortening network training and predicted time. Experiments show that the processing effect of this method on word segmentation is similar to that of BI-LSTM-CRF (Bidirectional- Long Short Term Memory-Conditional Random Field) model, but the average predicted processing speed is 1.94 times that of BI-LSTM-CRF, effectively improving the efficiency of word segmentation processing.

A Natural Language Processing Pipeline of Chinese Free-Text Radiology Reports for Liver Cancer Diagnosis

Despite the rapid development of natural language processing (NLP) implementation in electronic medical records (EMRs), Chinese EMRs processing remains challenging due to the limited corpus and specific grammatical characteristics, especially for radiology reports. In this study, we designed an NLP pipeline for the direct extraction of clinically relevant features from Chinese radiology reports, which is the first key step in computer-aided radiologic diagnosis. The pipeline was comprised of named entity recognition, synonyms normalization, and relationship extraction to finally derive the radiological features composed of one or more terms. In named entity recognition, we incorporated lexicon into deep learning model bidirectional long short-term memory-conditional random field (BiLSTM-CRF), and the model finally achieved an F1 score of 93.00%. With the extracted radiological features, least absolute shrinkage and selection operator and machine learning methods (support vector machine, random forest, decision tree, and logistic regression) were used to build the classifiers for liver cancer prediction. For liver cancer diagnosis, random forest had the highest predictive performance in liver cancer diagnosis (F1 score 86.97%, precision 87.71%, and recall 86.25%). This work was a comprehensive NLP study focusing on Chinese radiology reports and the application of NLP in cancer risk prediction. The proposed NLP pipeline for the radiological feature extraction could be easily implemented in other kinds of Chinese clinical texts and other disease predictive tasks.

Post Text Processing of Chinese Speech Recognition Based on Bidirectional LSTM Networks and CRF

With the rapid development of Internet of Things Technology, speech recognition has been applied more and more widely. Chinese Speech Recognition is a complex process. In the process of speech-to-text conversion, due to the influence of dialect, environmental noise, and context, the accuracy of speech-to-text in multi-round dialogues and specific contexts is still not high. After the general speech recognition technology, the text after speech recognition can be detected and corrected in the specific context, which is helpful to improve the robustness of text comprehension and is a beneficial supplement to the speech recognition technology. In this paper, a text processing model after Chinese Speech Recognition is proposed, which combines a bidirectional long short-term memory (LSTM) network with a conditional random field (CRF) model. The task is divided into two stages: text error detection and text error correction. In this paper, a bidirectional long short-term memory (Bi-LSTM) network and conditional random field are used in two stages of text error detection and text error correction respectively. Through verification and system test on the SIGHAN 2013 Chinese Spelling Check (CSC) dataset, the experimental results show that the model can effectively improve the accuracy of text after speech recognition.

Annotating and detecting phenotypic information for chronic obstructive pulmonary disease

ObjectivesChronic obstructive pulmonary disease (COPD) phenotypes cover a range of lung abnormalities. To allow text mining methods to identify pertinent and potentially complex information about these phenotypes from textual data, we have developed a novel annotated corpus, which we use to train a neural network-based named entity recognizer to detect fine-grained COPD phenotypic information.Materials and methodsSince COPD phenotype descriptions often mention other concepts within them (proteins, treatments, etc.), our corpus annotations include both outermost phenotype descriptions and concepts nested within them. Our neural layered bidirectional long short-term memory conditional random field (BiLSTM-CRF) network firstly recognizes nested mentions, which are fed into subsequent BiLSTM-CRF layers, to help to recognize enclosing phenotype mentions.ResultsOur corpus of 30 full papers (available at: http://www.nactem.ac.uk/COPD) is annotated by experts with 27 030 phenotype-related concept mentions, most of which are automatically linked to UMLS Metathesaurus concepts. When trained using the corpus, our BiLSTM-CRF network outperforms other popular approaches in recognizing detailed phenotypic information.DiscussionInformation extracted by our method can facilitate efficient location and exploration of detailed information about phenotypes, for example, those specifically concerning reactions to treatments.ConclusionThe importance of our corpus for developing methods to extract fine-grained information about COPD phenotypes is demonstrated through its successful use to train a layered BiLSTM-CRF network to extract phenotypic information at various levels of granularity. The minimal human intervention needed for training should permit ready adaption to extracting phenotypic information about other diseases.

A Character-Level BiLSTM-CRF Model With Multi-Representations for Chinese Event Detection

Using the word as a basic unit may undermine Chinese event detection model’s performance because of the inaccurate word boundaries generated by segmentation tools. Besides, word embeddings are contextual independent and cannot handle the polysemy of event triggers, which may prevent us from obtaining the desired performance. To address these issues, we propose a BiLSTM-CRF (Bidirectional Long Short-Term Memory Conditional Random Field) model using contextualized representations, which regards event detection task as a character-level sequence labeling problem and uses contextualized representations to disambiguate event triggers. Experiments show that our proposed method sets a new state-of-the-art, which proves Chinese characters could replace words for the Chinese event detection task. Besides, using contextualized representation reduces the false positive case, which verifies that this kind of representation could remedy the weakness of the word embedding technique. Based on the results, we believe that character-level models are worth exploring in the future.