Factoid Questions Research Articles

A Knowledge Graph Embedding Model for Answering Factoid Entity Questions

Factoid entity questions (FEQ), which seek answers in the form of a single entity from knowledge sources such as DBpedia and Wikidata, constitute a substantial portion of user queries in search engines. This paper introduces the Knowledge Graph Embedding model for Factoid Entity Question answering (KGE-FEQ). Leveraging a textual knowledge graph derived from extensive text collections, KGE-FEQ encodes textual relationships between entities. The model employs a two-step process: (1) Triple Retrieval, where relevant triples are retrieved from the textual knowledge graph based on semantic similarities to the question, and (2) Answer Selection, where a knowledge graph embedding approach is utilized for answering the question. This involves positioning the embedding for the answer entity close to the embedding of the question entity, incorporating a vector representing the question and textual relations between entities. Extensive experiments evaluate the performance of the proposed approach, comparing KGE-FEQ to state-of-the-art baselines in factoid entity question answering and the most advanced open-domain question answering techniques applied to FEQs. The results show that KGE-FEQ outperforms existing methods across different datasets. Ablation studies highlights the effectiveness of KGE-FEQ when both the question and textual relations between entities are considered for answering questions.

Semantic features analysis for biomedical lexical answer type prediction using ensemble learning approach

Lexical answer type prediction is integral to biomedical question–answering systems. LAT prediction aims to predict the expected answer’s semantic type of a factoid or list-type biomedical question. It also aids in the answer processing stage of a QA system to assign a high score to the most relevant answers. Although considerable research efforts exist for LAT prediction in diverse domains, it remains a challenging biomedical problem. LAT prediction for the biomedical field is a multi-label classification problem, as one biomedical question might have more than one expected answer type. Achieving high performance on this task is challenging as biomedical questions have limited lexical features. One biomedical question must be assigned multiple labels given these limited lexical features. In this paper, we develop a novel feature set (lexical, noun concepts, verb concepts, protein–protein interactions, and biomedical entities) from these lexical features. Using ensemble learning with bagging, we use the label power set transformation technique to classify multi-label. We evaluate the integrity of our proposed methodology on the publicly available multi-label biomedical questions dataset (MLBioMedLAT) and compare it with twelve state-of-the-art multi-label classification algorithms. Our proposed method attains a micro-F1 score of 77%, outperforming the baseline model by 25.5%.

A Unified Framework for Contextual and Factoid Question Generation

Question generation (QG) aims to automatically generate fluent and relevant questions, where the two most mainstream directions are generating questions from unstructured contextual texts (CQG), such as news articles, and generating questions from structured factoid texts (FQG), such as knowledge graphs or tables. Existing methods for these two tasks mainly face challenges of limited internal structural information as well as scarce background information, while these two tasks can benefit each other for alleviating these issues. For example, when meeting the entity mention “United Kingdom” in CQG, it can be inferred that it is a country in European continent based on the structural knowledge “(Europe, countries_within, United Kingdom)” in FQG. And when meeting the entity “Houston Rockets” in FQG, more background information, such as “an American professional basketball team based in Houston since 1971”, can be found in the related passages of CQG. To this end, we propose a unified framework for the tasks of CQG and FQG, where: (i) two types of task-sharing modules are developed to learn shared contextual and structural knowledge, where the task format is unified with a pseudo passage reformulation strategy; (ii) for the CQG task, a task-specific knowledge module with a knowledge selection and aggregation mechanism is introduced, so as to incorporate more factoid knowledge from external knowledge graphs and alleviate the word ambiguity problem; and (iii) for the FQG task, a task-specific passage module with a multi-level passage fusion mechanism is designed to extract fine-grained word-level knowledge. Experimental results in both automatic and human evaluation show the effectiveness of our proposed method.

SS-BERT: A Semantic Information Selecting Approach for Open-Domain Question Answering

Open-Domain Question Answering (Open-Domain QA) aims to answer any factoid questions from users. Recent progress in Open-Domain QA adopts the “retriever-reader” structure, which has proven effective. Retriever methods are mainly categorized as sparse retrievers and dense retrievers. In recent work, the dense retriever showed a stronger semantic interpretation than the sparse retriever. When training a dual-encoder dense retriever for document retrieval and reranking, there are two challenges: negative selection and a lack of training data. In this study, we make three major contributions to this topic: negative selection by query generation, data augmentation from negatives, and a passage evaluation method. We prove that the model performs better by focusing on false negatives and data augmentation in the Open-Domain QA passage rerank task. Our model outperforms other single dual-encoder rerankers over BERT-base and BM25 by 0.7 in MRR@10, achieving the highest Recall@50 and the max Recall@1000, which is restricted by the BM25 retrieval results.

Adversarial Knowledge Distillation Based Biomedical Factoid Question Answering.

Biomedical factoid question answering is an essential application for biomedical information sharing. Recently, neural network based approaches have shown remarkable performance for this task. However, due to the scarcity of annotated data which requires intensive knowledge of expertise, training a robust model on limited-scale biomedical datasets remains a challenge. Previous works solve this problem by introducing useful knowledge. It is found that the interaction between question and answer (QA-interaction) is also a kind of knowledge which could help extract answer accurately. This research develops a knowledge distillation framework for biomedical factoid question answering, in which a teacher model as the knowledge source of QA-interaction is designed to enhance the student model. In addition, to further alleviate the problem of limited-scale dataset, a novel adversarial knowledge distillation technique is proposed to robustly distill the knowledge from teacher model to student model by constructing perturbed examples as additional training data. By forcing the student model to mimic the predicted distributions of teacher model on both original examples and perturbed examples, the knowledge of QA-interaction can be learned by student model. We evaluate the proposed framework on the widely used BioASQ datasets, and experimental results have shown the proposed method's promising potential.

Hyperbolic Directed Hypergraph-Based Reasoning for Multi-Hop KBQA

The target of the multi-hop knowledge base question-answering task is to find answers of some factoid questions by reasoning across multiple knowledge triples in the knowledge base. Most of the existing methods for multi-hop knowledge base question answering based on a general knowledge graph ignore the semantic relationship between each hop. However, modeling the knowledge base as a directed hypergraph has the problems of sparse incidence matrices and asymmetric Laplacian matrices. To make up for the deficiency, we propose a directed hypergraph convolutional network modeled on hyperbolic space, which can better deal with the sparse structure, and effectively adapt to the problem of an asymmetric incidence matrix of directed hypergraphs modeled on a knowledge base. We propose an interpretable KBQA model based on the hyperbolic directed hypergraph convolutional neural network named HDH-GCN which can update relation semantic information hop-by-hop and pays attention to different relations at different hops. The model can improve the accuracy of the multi-hop knowledge base question-answering task, and has application value in text question answering, human–computer interactions and other fields. Extensive experiments on benchmarks—PQL, MetaQA—demonstrate the effectiveness and universality of our HDH-GCN model, leading to state-of-the-art performance.

Template-Based Question Answering System Over the Semantic Web

A question answering system is the most promising way of retrieving data from the available knowledge base to the end-users to get the appropriate result for their questions. Many question answering systems convert the questions into triples which are mapped to the knowledge base from which the answer is derived. However, these triples do not express the semantic representation of the question, due to which the answers cannot be located. To handle this, a template-based approach is proposed that classifies the question types and finds appropriate SPARQL query template for each type including comparatives and superlatives. The SPARQL query built is executed in the DBpedia endpoint and results are obtained. Compared with other factoid question answering systems, the proposed approach has the potential to deal with a large number of question types, including comparatives and superlatives. Also, the experimental evaluations of the system performed on the QALD 8 dataset presents good performance and can help users to find answers to their questions.

Direct relation detection for knowledge-based question answering

This study addresses the problem of relation detection for answering single-relation factoid questions over knowledge bases (KBs). In this kind of questions, the answer is obtained from a single KB fact in the form of subject-predicate-object. Conventional fact extraction methods have two steps: entity linking and relation detection, in which the output of the entity linking is used by the relation detection step to first find candidate relations, and then choose the best relation from candidate relations. Such methods have difficulties with the relation detection if there is an error or ambiguity in the entity linking step. This paper explores the relation detection task without the entity-linking step utilizing the hierarchical structure of relations and an out-of-box POS tagger. As relations are of different levels of abstraction, the proposed solution uses multiple classifiers in pipeline, each of which uses separate BiGRU neural networks fed with questions embedded with one-hot encoding at the character level. Besides, to increase the accuracy of the proposed model and to avoid the need for large amounts of training data, after each word of the question, its POS tag is inserted before feeding the network. The experimental results show that the accuracy of the proposed solution for the direct relation detection is 89.5%. In addition, the proposed solution can be used for the indirect relation detection whose accuracy is 96.3%, which is higher than state-of-the-art relation detection techniques. Finally, the positive effects of using POS tags have been examined.

Template based Question Answering System Over Semantic Web

Question Answering system is the most promising way of retrieving data from the available knowledge base to the end-users, to get the appropriate result for their questions. Many Question Answering systems convert the questions into triples which are mapped to the Knowledge base from which answer is derived. However, these triples do not express the semantic representation of the question, due to which the answers cannot be located. To handle this, a template-based approach is proposed which classifies the question types and finds appropriate SPARQL query template for each type including comparatives and superlatives. The SPARQL query built is executed in the DBpedia endpoint and results are obtained. Compared with other factoid question answering systems, the proposed approach has the potential to deal with a large number of question types, including comparatives and superlatives. Also, the experimental evaluations of the system performed on the QALD 8 dataset presents good performance and can help users to find answers to their questions.

SentiMedQAer: A Transfer Learning-Based Sentiment-Aware Model for Biomedical Question Answering.

Recent advances have witnessed a trending application of transfer learning in a broad spectrum of natural language processing (NLP) tasks, including question answering (QA). Transfer learning allows a model to inherit domain knowledge obtained from an existing model that has been sufficiently pre-trained. In the biomedical field, most QA datasets are limited by insufficient training examples and the presence of factoid questions. This study proposes a transfer learning-based sentiment-aware model, named SentiMedQAer, for biomedical QA. The proposed method consists of a learning pipeline that utilizes BioBERT to encode text tokens with contextual and domain-specific embeddings, fine-tunes Text-to-Text Transfer Transformer (T5), and RoBERTa models to integrate sentiment information into the model, and trains an XGBoost classifier to output a confidence score to determine the final answer to the question. We validate SentiMedQAer on PubMedQA, a biomedical QA dataset with reasoning-required yes/no questions. Results show that our method outperforms the SOTA by 15.83% and a single human annotator by 5.91%.

Evidence Integration for Multi-Hop Reading Comprehension With Graph Neural Networks

Multi-hop reading comprehension focuses on one type of factoid question, where a system needs to properly integrate multiple pieces of evidence to correctly answer a question. Previous work approximates global evidence with local coreference information, encoding coreference chains with DAG-styled GRU layers within a gated-attention reader. However, coreference is limited in providing information for rich inference. We introduce a new method for better connecting global evidence, which forms more complex graphs compared to DAGs. To perform evidence integration on our graphs, we investigate two recent graph neural networks, namely graph convolutional network (GCN) and graph recurrent network (GRN). Experiments on two standard datasets show that richer global information leads to better answers. Our approach shows highly competitive performances on these datasets without deep language models (such as ELMo).

Generating Factoid Questions with Question Type Enhanced Representation and Attention-based Copy Mechanism

Question generation over knowledge bases is an important research topic. How to deal with rare and low-frequency words in traditional generation models is a key challenge for question generation. Although the copy mechanism provides significant performance improvements, the original copy mechanism weakens the focus on aspect generation in the overall representations. In this article, we present a novel method to improve question generation with a question type enhanced representation and attention-based copy mechanism. The proposed method exploits the advantages of the generate mode in the copy mechanism and replaces objects in the factual triples with question types, which attempts to improve the output quality in the generate mode and effectively generate questions with proper interrogative words. We evaluate the proposed method on two standard benchmark datasets. The experimental results demonstrate that our proposed method can produce higher-quality questions than these of the Encoder-Decoder-based and CopyNet-based methods.

Neural factoid geospatial question answering

Existing question answering systems struggle to answer factoid questions when geospatial information is involved. This is because most systems cannot accurately detect the geospatial semantic elements from the natural language questions, or capture the semantic relationships between those elements. In this paper, we propose a geospatial semantic encoding schema and a semantic graph representation which captures the semantic relations and dependencies in geospatial questions. We demonstrate that our proposed graph representation approach aids in the translation from natural language to a formal, executable expression in a query language. To decrease the need for people to provide explanatory information as part of their question and make the translation fully automatic, we treat the semantic encoding of the question as a sequential tagging task, and the graph generation of the query as a semantic dependency parsing task. We apply neural network approaches to automatically encode the geospatial questions into spatial semantic graph representations. Compared with current template-based approaches, our method generalises to a broader range of questions, including those with complex syntax and semantics. Our proposed approach achieves better results on GeoData201 than existing methods.

VQA as a factoid question answering problem: A novel approach for knowledge-aware and explainable visual question answering

With recent advancements in machine perception and scene understanding, Visual Question Answering (VQA) has garnered much attraction from researchers in the direction of training neural models for jointly analyzing, grounding and reasoning over the multi-modal space of image visual context and natural language in order to answer natural language questions pertaining to the image contents. However, though recent works have achieved significant improvement over state-of-art models for answering questions that are answerable by solely referring to the visual context of the image, such models are often limited, being incapable of tackling questions involving external world knowledge beyond the visible contents. Though recently, research has been driven towards tackling external knowledge based VQA as well, there is significant room for improvement as limited studies exist in this area.Inspired by the aforementioned challenges involved, this paper is aimed at answering free form and open ended natural language questions, not limited to visual context of an image, but external world knowledge as well. With this motive, inspired by human cognitive abilities of comprehending and reasoning answers when given a set of facts, this paper proposes a novel model architecture to model VQA as a factoid question answering problem, leveraging state-of-the-art deep learning techniques for reasoning and inferring answers to free form questions, in an attempt of improving the state-of-art in open ended visual question answering.

Indonesian Question Answering System for Factoid Questions using Face Beauty Products Knowledge Graph

Question answering (QA) system is developed to find the right answers from natural language questions. QA systems can be used for building chatbots or even search engines. In this study, we’ve built an Indonesian QA system that uses Anindya Knowledge Graph as its data source. The idea behind this QA system is translating questions into SPARQL queries. The proposed solution consists of four modules, namely question classification, information extraction, token mapping, and query construction. The question classification and the information extraction modules were experimented using SVM, LSTM, and fine-tuning IndoBERT. The text representations were also tested to find the best result among tf-idf, FastText, and IndoBERT. In our experiment, we found that the fine-tuning IndoBERT model had obtained the best performance on both question classification and information extraction modules.

A survey on non-factoid question answering systems

Question Answering System (QAS) aims at providing the most appropriate answer to the user’s question asked in any natural language. It emerged as a future of web search that crawls structured and unstructured collections of documents to find answers. A lot of research has already been done on answering factoid questions and achieved promising accuracy, whereas non-factoid question answering is a challenging task and explores newer avenues for fact-finding and analysis. The research on handling different modules of non-factoid question answering has drastically increased from 2012 which evokes the need for summarizing and providing a comprehensive analysis of the work that has already been done in this field. The survey identifies different challenges faced in each module, how they have been addressed by the researchers with their advantages or disadvantages and further improvements. The survey is written with an objective to help the researchers understand the field with how it has been undertaken by different approaches with their impact on performance. The paper finally concludes by investigating research gaps and research objectives that are promising to get an insight into future research directions.

Analyzing Linguistic Features for Classifying Why-Type Non-Factoid Questions

Why-type non-factoid questions are complex and difficult to answer compared to factoid questions. A challenge in finding an accurate answer to a non-factoid question is to understand the intent of user as it differs with their knowledge and also the context of the question in which it is being asked. Predicting correct type of a question and its answer by a classification model is an important issue as it affects the subsequent processing of its answer. In this paper, a classification model is proposed which is trained by a combination of lexical, syntactic, and semantic features to classify open-domain why-type questions. Various supervised classifiers are trained on a featured dataset out of which support vector machine achieves the highest accuracy of 81% in determining question type and 76.8% in determining answer type which shows 14.6% improvement in predicting an answer type over a baseline why-type classifier with 62.2% accuracy.

External features enriched model for biomedical question answering

BackgroundBiomedical question answering (QA) is a sub-task of natural language processing in a specific domain, which aims to answer a question in the biomedical field based on one or more related passages and can provide people with accurate healthcare-related information. Recently, a lot of approaches based on the neural network and large scale pre-trained language model have largely improved its performance. However, considering the lexical characteristics of biomedical corpus and its small scale dataset, there is still much improvement room for biomedical QA tasks.ResultsInspired by the importance of syntactic and lexical features in the biomedical corpus, we proposed a new framework to extract external features, such as part-of-speech and named-entity recognition, and fused them with the original text representation encoded by pre-trained language model, to enhance the biomedical question answering performance. Our model achieves an overall improvement of all three metrics on BioASQ 6b, 7b, and 8b factoid question answering tasks.ConclusionsThe experiments on BioASQ question answering dataset demonstrated the effectiveness of our external feature-enriched framework. It is proven by the experiments conducted that external lexical and syntactic features can improve Pre-trained Language Model’s performance in biomedical domain question answering task.

Named Entity Aware Transfer Learning for Biomedical Factoid Question Answering.

Biomedical factoid question answering is an important task in biomedical question answering applications. It has attracted much attention because of its reliability. In question answering systems, better representation of words is of great importance, and proper word embedding can significantly improve the performance of the system. With the success of pretrained models in general natural language processing tasks, pretrained models have been widely used in biomedical areas, and many pretrained model-based approaches have been proven effective in biomedical question-answering tasks. In addition to proper word embedding, name entities also provide important information for biomedical question answering. Inspired by the concept of transfer learning, in this study, we developed a mechanism to fine-tune BioBERT with a named entity dataset to improve the question answering performance. Furthermore, we applied BiLSTM to encode the question text to obtain sentence-level information. To better combine the question level and token level information, we use bagging to further improve the overall performance. The proposed framework was evaluated on BioASQ 6b and 7b datasets, and the results have shown that our proposed framework can outperform all baselines.

Revealing Opinions for COVID-19 Questions Using a Context Retriever, Opinion Aggregator, and Question-Answering Model: Model Development Study.

BackgroundCOVID-19 has challenged global public health because it is highly contagious and can be lethal. Numerous ongoing and recently published studies about the disease have emerged. However, the research regarding COVID-19 is largely ongoing and inconclusive.ObjectiveA potential way to accelerate COVID-19 research is to use existing information gleaned from research into other viruses that belong to the coronavirus family. Our objective is to develop a natural language processing method for answering factoid questions related to COVID-19 using published articles as knowledge sources.MethodsGiven a question, first, a BM25-based context retriever model is implemented to select the most relevant passages from previously published articles. Second, for each selected context passage, an answer is obtained using a pretrained bidirectional encoder representations from transformers (BERT) question-answering model. Third, an opinion aggregator, which is a combination of a biterm topic model and k-means clustering, is applied to the task of aggregating all answers into several opinions.ResultsWe applied the proposed pipeline to extract answers, opinions, and the most frequent words related to six questions from the COVID-19 Open Research Dataset Challenge. By showing the longitudinal distributions of the opinions, we uncovered the trends of opinions and popular words in the articles published in the five time periods assessed: before 1990, 1990-1999, 2000-2009, 2010-2018, and since 2019. The changes in opinions and popular words agree with several distinct characteristics and challenges of COVID-19, including a higher risk for senior people and people with pre-existing medical conditions; high contagion and rapid transmission; and a more urgent need for screening and testing. The opinions and popular words also provide additional insights for the COVID-19–related questions.ConclusionsCompared with other methods of literature retrieval and answer generation, opinion aggregation using our method leads to more interpretable, robust, and comprehensive question-specific literature reviews. The results demonstrate the usefulness of the proposed method in answering COVID-19–related questions with main opinions and capturing the trends of research about COVID-19 and other relevant strains of coronavirus in recent years.