Foundation models are reshaping artificial intelligence,yet their deployment in specialised domains such as agricultural question answering(AQA)still faces challenges including data scarcity and barriers to domainspec...Foundation models are reshaping artificial intelligence,yet their deployment in specialised domains such as agricultural question answering(AQA)still faces challenges including data scarcity and barriers to domainspecific knowledge.To systematically review recent progress in this area,this paper adopts a task–paradigmperspective and examines applications across three major AQA task families.For text-based QA,we analyse the strengths and limitations of retrieval-based,generative,and hybrid approaches built on large languagemodels,revealing a clear trend toward hybrid paradigms that balance precision and flexibility.For visual diagnosis,we discuss techniques such as crossmodal alignment and prompt-driven generation,which are pushing systems beyond simple pest and disease recognition toward deeper causal reasoning.Formultimodal reasoning,we show how the fusion of heterogeneous data—including text,images,speech,and sensor streams—enables comprehensive decision-making for diagnosis,monitoring,and yield prediction.To address the lack of unified benchmarks,we further propose a standardised evaluation protocol and a diagnostic taxonomy specifically designed to characterise agriculture-specific errors.Finally,we outline a concreteAQA roadmap that emphasises safety alignment,hallucination control,and lightweight deployment,aiming to guide future systems toward greater efficiency,trustworthiness,and sustainability.展开更多
Background External knowledge representations play an essential role in knowledge-based visual question and answering to better understand complex scenarios in the open world.Recent entity-relationship embedding appro...Background External knowledge representations play an essential role in knowledge-based visual question and answering to better understand complex scenarios in the open world.Recent entity-relationship embedding approaches are deficient in representing some complex relations,resulting in a lack of topic-related knowledge and redundancy in topic-irrelevant information.Methods To this end,we propose MKEAH:Multimodal Knowledge Extraction and Accumulation on Hyperplanes.To ensure that the lengths of the feature vectors projected onto the hyperplane compare equally and to filter out sufficient topic-irrelevant information,two losses are proposed to learn the triplet representations from the complementary views:range loss and orthogonal loss.To interpret the capability of extracting topic-related knowledge,we present the Topic Similarity(TS)between topic and entity-relations.Results Experimental results demonstrate the effectiveness of hyperplane embedding for knowledge representation in knowledge-based visual question answering.Our model outperformed state-of-the-art methods by 2.12%and 3.24%on two challenging knowledge-request datasets:OK-VQA and KRVQA,respectively.Conclusions The obvious advantages of our model in TS show that using hyperplane embedding to represent multimodal knowledge can improve its ability to extract topic-related knowledge.展开更多
Knowledge-based VisualQuestion Answering(VQA)requires the integration of visual information with external knowledge reasoning.Existing approaches typically retrieve information from external corpora and rely on pretra...Knowledge-based VisualQuestion Answering(VQA)requires the integration of visual information with external knowledge reasoning.Existing approaches typically retrieve information from external corpora and rely on pretrained language models for reasoning.However,their performance is often hindered by the limited capabilities of retrievers and the constrained size of knowledge bases.Moreover,relying on image captions to bridge the modal gap between visual and language modalities can lead to the omission of critical visual details.To address these limitations,we propose the Reflective Chain-of-Thought(ReCoT)method,a simple yet effective framework inspired by metacognition theory.ReCoT effectively activates the reasoning capabilities ofMultimodal Large LanguageModels(MLLMs),providing essential visual and knowledge cues required to solve complex visual questions.It simulates a metacognitive reasoning process that encompasses monitoring,reflection,and correction.Specifically,in the initial generation stage,an MLLM produces a preliminary answer that serves as the model’s initial cognitive output.During the reflective reasoning stage,this answer is critically examined to generate a reflective rationale that integrates key visual evidence and relevant knowledge.In the final refinement stage,a smaller language model leverages this rationale to revise the initial prediction,resulting in amore accurate final answer.By harnessing the strengths ofMLLMs in visual and knowledge grounding,ReCoT enables smaller language models to reason effectively without dependence on image captions or external knowledge bases.Experimental results demonstrate that ReCoT achieves substantial performance improvements,outperforming state-of-the-art methods by 2.26%on OK-VQA and 5.8%on A-OKVQA.展开更多
Medical visual question answering(MedVQA)faces unique challenges due to the high precision required for images and the specialized nature of the questions.These challenges include insufficient feature extraction capab...Medical visual question answering(MedVQA)faces unique challenges due to the high precision required for images and the specialized nature of the questions.These challenges include insufficient feature extraction capabilities,a lack of textual priors,and incomplete information fusion and interaction.This paper proposes an enhanced bootstrapping language-image pre-training(BLIP)model for MedVQA based on multimodal feature augmentation and triple-path collaborative attention(FCA-BLIP)to address these issues.First,FCA-BLIP employs a unified bootstrap multimodal model architecture that integrates ResNet and bidirectional encoder representations from Transformer(BERT)models to enhance feature extraction capabilities.It enables a more precise analysis of the details in images and questions.Next,the pre-trained BLIP model is used to extract features from image-text sample pairs.The model can understand the semantic relationships and shared information between images and text.Finally,a novel attention structure is developed to fuse the multimodal feature vectors,thereby improving the alignment accuracy between modalities.Experimental results demonstrate that the proposed method performs well in clinical visual question-answering tasks.For the MedVQA task of staging diabetic macular edema in fundus imaging,the proposed method outperforms the existing major models in several performance metrics.展开更多
Aiming at the lack of professional knowledge to guide apparel recommendation,an apparel recommendation method based on image design expert knowledge has been proposed.Then,apparel recommendation knowledge graphs have ...Aiming at the lack of professional knowledge to guide apparel recommendation,an apparel recommendation method based on image design expert knowledge has been proposed.Then,apparel recommendation knowledge graphs have been created and a apparel recommendation question and answer(Q&A)system has been designed and implemented.The question templates in the apparel recommendation domain were defined,the task of recognizing the named entities of question sentences was completed by the Bi-directional encoder representations from transformer-Bi-directional long short-term memory-conditional random field(BERT-BiLSTM-CRF)model,and the question template with the highest matching degree to the user’s question was obtained by using term frequency-inverse document frequency(TF-IDF)algorithm.The corresponding cypher graph database query statement was generated to retrieve the knowledge graph for answers,and iFLYTEK’s voice application programming interface(API)was called to implement the Q&A.The experimental results have shown that the Q&A system has a high accuracy rate and application value in the field of apparel recommendations.展开更多
To improve question answering （QA） performance based on real-world web data sets,a new set of question classes and a general answer re-ranking model are defined.With pre-defined dictionary and grammatical analysis,t...To improve question answering （QA） performance based on real-world web data sets,a new set of question classes and a general answer re-ranking model are defined.With pre-defined dictionary and grammatical analysis,the question classifier draws both semantic and grammatical information into information retrieval and machine learning methods in the form of various training features,including the question word,the main verb of the question,the dependency structure,the position of the main auxiliary verb,the main noun of the question,the top hypernym of the main noun,etc.Then the QA query results are re-ranked by question class information.Experiments show that the questions in real-world web data sets can be accurately classified by the classifier,and the QA results after re-ranking can be obviously improved.It is proved that with both semantic and grammatical information,applications such as QA, built upon real-world web data sets, can be improved,thus showing better performance.展开更多
Knowledge-based visual question answering(KB-VQA),requiring external world knowledge beyond the image for reasoning,is more challenging than traditional visual question answering.Recent works have demonstrated the eff...Knowledge-based visual question answering(KB-VQA),requiring external world knowledge beyond the image for reasoning,is more challenging than traditional visual question answering.Recent works have demonstrated the effectiveness of using a large(vision)language model as an implicit knowledge source to acquire the necessary information.However,the knowledge stored in large models(LMs)is often coarse-grained and inaccurate,causing questions requiring finer-grained information to be answered incorrectly.In this work,we propose a variational expectation-maximization(EM)framework that bootstraps the VQA performance of LMs with its own answer.In contrast to former VQA pipelines,we treat the outside knowledge as a latent variable.In the E-step,we approximate the posterior with two components:First,a rough answer,e.g.,a general description of the image,which is usually the strength of LMs,and second,a multi-modal neural retriever to retrieve question-specific knowledge from an external knowledge base.In the M-step,the training objective optimizes the ability of the original LMs to generate rough answers as well as refined answers based on the retrieved information.Extensive experiments show that our proposed framework,BootLM,has a strong retrieval ability and achieves state-of-the-art performance on knowledge-based VQA tasks.展开更多
Based on personal observation and students’ self-reports, this article studies the characteristics presented by students in answering questions. It is found that Chinese culture plays an important part in explaining ...Based on personal observation and students’ self-reports, this article studies the characteristics presented by students in answering questions. It is found that Chinese culture plays an important part in explaining their specific behaviors. Hence, in order to lessen the negative effect of it, the author offers her own suggestions.展开更多
A passage retrieval strategy for web-based question answering （QA） systems is proposed in our QA system. It firstly analyzes the question based on semantic patterns to obtain its syntactic and semantic information a...A passage retrieval strategy for web-based question answering （QA） systems is proposed in our QA system. It firstly analyzes the question based on semantic patterns to obtain its syntactic and semantic information and then form initial queries. The queries are used to retrieve documents from the World Wide Web （WWW） using the Google search engine. The queries are then rewritten to form queries for passage retrieval in order to improve the precision. The relations between keywords in the question are employed in our query rewrite method. The experimental result on the question set of the TREC-2003 passage task shows that our system performs well for factoid questions.展开更多
Over the last couple of decades,community question-answering sites(CQAs)have been a topic of much academic interest.Scholars have often leveraged traditional machine learning(ML)and deep learning(DL)to explore the eve...Over the last couple of decades,community question-answering sites(CQAs)have been a topic of much academic interest.Scholars have often leveraged traditional machine learning(ML)and deep learning(DL)to explore the ever-growing volume of content that CQAs engender.To clarify the current state of the CQA literature that has used ML and DL,this paper reports a systematic literature review.The goal is to summarise and synthesise the major themes of CQA research related to(i)questions,(ii)answers and(iii)users.The final review included 133 articles.Dominant research themes include question quality,answer quality,and expert identification.In terms of dataset,some of the most widely studied platforms include Yahoo!Answers,Stack Exchange and Stack Overflow.The scope of most articles was confined to just one platform with few cross-platform investigations.Articles with ML outnumber those with DL.Nonetheless,the use of DL in CQA research is on an upward trajectory.A number of research directions are proposed.展开更多
Deep learning models have been shown to have great advantages in answer selection tasks.The existing models,which employ encoder-decoder recurrent neural network(RNN),have been demonstrated to be effective.However,the...Deep learning models have been shown to have great advantages in answer selection tasks.The existing models,which employ encoder-decoder recurrent neural network(RNN),have been demonstrated to be effective.However,the traditional RNN-based models still suffer from limitations such as 1)high-dimensional data representation in natural language processing and 2)biased attentive weights for subsequent words in traditional time series models.In this study,a new answer selection model is proposed based on the Bidirectional Long Short-Term Memory(Bi-LSTM)and attention mechanism.The proposed model is able to generate the more effective question-answer pair representation.Experiments on a question answering dataset that includes information from multiple fields show the great advantages of our proposed model.Specifically,we achieve a maximum improvement of 3.8%over the classical LSTM model in terms of mean average precision.展开更多
In Chinese question answering system, because there is more semantic relation in questions than that in query words, the precision can be improved by expanding query while using natural language questions to retrieve ...In Chinese question answering system, because there is more semantic relation in questions than that in query words, the precision can be improved by expanding query while using natural language questions to retrieve documents. This paper proposes a new approach to query expansion based on semantics and statistics Firstly automatic relevance feedback method is used to generate a candidate expansion word set. Then the expanded query words are selected from the set based on the semantic similarity and seman- tic relevancy between the candidate words and the original words. Experiments show the new approach is effective for Web retrieval and out-performs the conventional expansion approaches.展开更多
Question answering systems offer a friendly interface for human beings to interact with massive online information. It is time consuming for users to retrieve useful medical information with search engines among massi...Question answering systems offer a friendly interface for human beings to interact with massive online information. It is time consuming for users to retrieve useful medical information with search engines among massive online websites. An effort is made to build a Chinese Question Answering System in Medical Domain(CQASMD) to provide useful medical information for users. A large medical knowledge base with more than 300 thousand medical terms and their descriptions is firstly constructed to store the structured medical knowledge data, and classified with the FastText model. Furthermore, a Word2Vec model is adopted to capture the semantic meanings of words, and the questions and answers are processed with sentence embedding to capture semantic context information. Users' questions are firstly classified and processed into a sentence vector and a matching algorithm is adopted to match the most similar question. After querying the constructed medical knowledge base, the corresponding answers to previous questions are responded to users. The architecture and flowchart of CQASMD is proposed, which will play an important role in self disease diagnosis and treatment.展开更多
Aiming at the relation linking task for question answering over knowledge base,especially the multi relation linking task for complex questions,a relation linking approach based on the multi-attention recurrent neural...Aiming at the relation linking task for question answering over knowledge base,especially the multi relation linking task for complex questions,a relation linking approach based on the multi-attention recurrent neural network(RNN)model is proposed,which works for both simple and complex questions.First,the vector representations of questions are learned by the bidirectional long short-term memory(Bi-LSTM)model at the word and character levels,and named entities in questions are labeled by the conditional random field(CRF)model.Candidate entities are generated based on a dictionary,the disambiguation of candidate entities is realized based on predefined rules,and named entities mentioned in questions are linked to entities in knowledge base.Next,questions are classified into simple or complex questions by the machine learning method.Starting from the identified entities,for simple questions,one-hop relations are collected in the knowledge base as candidate relations;for complex questions,two-hop relations are collected as candidates.Finally,the multi-attention Bi-LSTM model is used to encode questions and candidate relations,compare their similarity,and return the candidate relation with the highest similarity as the result of relation linking.It is worth noting that the Bi-LSTM model with one attentions is adopted for simple questions,and the Bi-LSTM model with two attentions is adopted for complex questions.The experimental results show that,based on the effective entity linking method,the Bi-LSTM model with the attention mechanism improves the relation linking effectiveness of both simple and complex questions,which outperforms the existing relation linking methods based on graph algorithm or linguistics understanding.展开更多
ExpertRecommendation(ER)aims to identify domain experts with high expertise and willingness to provide answers to questions in Community Question Answering(CQA)web services.How to model questions and users in the hete...ExpertRecommendation(ER)aims to identify domain experts with high expertise and willingness to provide answers to questions in Community Question Answering(CQA)web services.How to model questions and users in the heterogeneous content network is critical to this task.Most traditional methods focus on modeling questions and users based on the textual content left in the community while ignoring the structural properties of heterogeneous CQA networks and always suffering from textual data sparsity issues.Recent approaches take advantage of structural proximities between nodes and attempt to fuse the textual content of nodes for modeling.However,they often fail to distinguish the nodes’personalized preferences and only consider the textual content of a part of the nodes in network embedding learning,while ignoring the semantic relevance of nodes.In this paper,we propose a novel framework that jointly considers the structural proximity relations and textual semantic relevance to model users and questions more comprehensively.Specifically,we learn topology-based embeddings through a hierarchical attentive network learning strategy,in which the proximity information and the personalized preference of nodes are encoded and preserved.Meanwhile,we utilize the node’s textual content and the text correlation between adjacent nodes to build the content-based embedding through a meta-context-aware skip-gram model.In addition,the user’s relative answer quality is incorporated to promote the ranking performance.Experimental results show that our proposed framework consistently and significantly outperforms the state-of-the-art baselines on three real-world datasets by taking the deep semantic understanding and structural feature learning together.The performance of the proposed work is analyzed in terms of MRR,P@K,and MAP and is proven to be more advanced than the existing methodologies.展开更多
With the warming up and continuous development of machine learning,especially deep learning,the research on visual question answering field has made significant progress,with important theoretical research significanc...With the warming up and continuous development of machine learning,especially deep learning,the research on visual question answering field has made significant progress,with important theoretical research significance and practical application value.Therefore,it is necessary to summarize the current research and provide some reference for researchers in this field.This article conducted a detailed and in-depth analysis and summarized of relevant research and typical methods of visual question answering field.First,relevant background knowledge about VQA(Visual Question Answering)was introduced.Secondly,the issues and challenges of visual question answering were discussed,and at the same time,some promising discussion on the particular methodologies was given.Thirdly,the key sub-problems affecting visual question answering were summarized and analyzed.Then,the current commonly used data sets and evaluation indicators were summarized.Next,in view of the popular algorithms and models in VQA research,comparison of the algorithms and models was summarized and listed.Finally,the future development trend and conclusion of visual question answering were prospected.展开更多
In this work, a best answer recommendation model is proposed for a Question Answering (QA) system. A Community Question Answering System was subsequently developed based on the model. The system applies Brouwer Fixed ...In this work, a best answer recommendation model is proposed for a Question Answering (QA) system. A Community Question Answering System was subsequently developed based on the model. The system applies Brouwer Fixed Point Theorem to prove the existence of the desired voter scoring function and Normalized Google Distance (NGD) to show closeness between words before an answer is suggested to users. Answers are ranked according to their Fixed-Point Score (FPS) for each question. Thereafter, the highest scored answer is chosen as the FPS Best Answer (BA). For each question asked by user, the system applies NGD to check if similar or related questions with the best answer had been asked and stored in the database. When similar or related questions with the best answer are not found in the database, Brouwer Fixed point is used to calculate the best answer from the pool of answers on a question then the best answer is stored in the NGD data-table for recommendation purpose. The system was implemented using PHP scripting language, MySQL for database management, JQuery, and Apache. The system was evaluated using standard metrics: Reciprocal Rank, Mean Reciprocal Rank (MRR) and Discounted Cumulative Gain (DCG). The system eliminated longer waiting time faced by askers in a community question answering system. The developed system can be used for research and learning purposes.展开更多
Recently,pre-trained language representation models such as bidirec-tional encoder representations from transformers(BERT)have been performing well in commonsense question answering(CSQA).However,there is a problem th...Recently,pre-trained language representation models such as bidirec-tional encoder representations from transformers(BERT)have been performing well in commonsense question answering(CSQA).However,there is a problem that the models do not directly use explicit information of knowledge sources existing outside.To augment this,additional methods such as knowledge-aware graph network(KagNet)and multi-hop graph relation network(MHGRN)have been proposed.In this study,we propose to use the latest pre-trained language model a lite bidirectional encoder representations from transformers(ALBERT)with knowledge graph information extraction technique.We also propose to applying the novel method,schema graph expansion to recent language models.Then,we analyze the effect of applying knowledge graph-based knowledge extraction techniques to recent pre-trained language models and confirm that schema graph expansion is effective in some extent.Furthermore,we show that our proposed model can achieve better performance than existing KagNet and MHGRN models in CommonsenseQA dataset.展开更多
The original intention of visual question answering(VQA)models is to infer the answer based on the relevant information of the question text in the visual image,but many VQA models often yield answers that are biased ...The original intention of visual question answering(VQA)models is to infer the answer based on the relevant information of the question text in the visual image,but many VQA models often yield answers that are biased by some prior knowledge,especially the language priors.This paper proposes a mitigation model called language priors mitigation-VQA(LPM-VQA)for the language priors problem in VQA model,which divides language priors into positive and negative language priors.Different network branches are used to capture and process the different priors to achieve the purpose of mitigating language priors.A dynamically-changing language prior feedback objective function is designed with the intermediate results of some modules in the VQA model.The weight of the loss value for each answer is dynamically set according to the strength of its language priors to balance its proportion in the total VQA loss to further mitigate the language priors.This model does not depend on the baseline VQA architectures and can be configured like a plug-in to improve the performance of the model over most existing VQA models.The experimental results show that the proposed model is general and effective,achieving state-of-the-art accuracy in the VQA-CP v2 dataset.展开更多
In the context of power generation companies, vast amounts of specialized data and expert knowledge have been accumulated. However, challenges such as data silos and fragmented knowledge hinder the effective utilizati...In the context of power generation companies, vast amounts of specialized data and expert knowledge have been accumulated. However, challenges such as data silos and fragmented knowledge hinder the effective utilization of this information. This study proposes a novel framework for intelligent Question-and-Answer (Q&A) systems based on Retrieval-Augmented Generation (RAG) to address these issues. The system efficiently acquires domain-specific knowledge by leveraging external databases, including Relational Databases (RDBs) and graph databases, without additional fine-tuning for Large Language Models (LLMs). Crucially, the framework integrates a Dynamic Knowledge Base Updating Mechanism (DKBUM) and a Weighted Context-Aware Similarity (WCAS) method to enhance retrieval accuracy and mitigate inherent limitations of LLMs, such as hallucinations and lack of specialization. Additionally, the proposed DKBUM dynamically adjusts knowledge weights within the database, ensuring that the most recent and relevant information is utilized, while WCAS refines the alignment between queries and knowledge items by enhanced context understanding. Experimental validation demonstrates that the system can generate timely, accurate, and context-sensitive responses, making it a robust solution for managing complex business logic in specialized industries.展开更多
基金supported by the Ningxia Natural Science Foundation(2025AAC050001)the Scientific Research Startup Project for Full-Time Introduced High-Level Talents in Ningxia(2024BEH04130)+2 种基金the National Natural Science Foundation of China(32460444)the Ningxia Hui Autonomous Region Key Research and Development Program(2024BBF0101302,2023BDE02001)Supported by the Special Fund for Basic Research Business of Central Universities of North Minzu University(2025BG234,2023ZRLG12).
文摘Foundation models are reshaping artificial intelligence,yet their deployment in specialised domains such as agricultural question answering(AQA)still faces challenges including data scarcity and barriers to domainspecific knowledge.To systematically review recent progress in this area,this paper adopts a task–paradigmperspective and examines applications across three major AQA task families.For text-based QA,we analyse the strengths and limitations of retrieval-based,generative,and hybrid approaches built on large languagemodels,revealing a clear trend toward hybrid paradigms that balance precision and flexibility.For visual diagnosis,we discuss techniques such as crossmodal alignment and prompt-driven generation,which are pushing systems beyond simple pest and disease recognition toward deeper causal reasoning.Formultimodal reasoning,we show how the fusion of heterogeneous data—including text,images,speech,and sensor streams—enables comprehensive decision-making for diagnosis,monitoring,and yield prediction.To address the lack of unified benchmarks,we further propose a standardised evaluation protocol and a diagnostic taxonomy specifically designed to characterise agriculture-specific errors.Finally,we outline a concreteAQA roadmap that emphasises safety alignment,hallucination control,and lightweight deployment,aiming to guide future systems toward greater efficiency,trustworthiness,and sustainability.
基金Supported by National Nature Science Foudation of China(61976160,61906137,61976158,62076184,62076182)Shanghai Science and Technology Plan Project(21DZ1204800)。
文摘Background External knowledge representations play an essential role in knowledge-based visual question and answering to better understand complex scenarios in the open world.Recent entity-relationship embedding approaches are deficient in representing some complex relations,resulting in a lack of topic-related knowledge and redundancy in topic-irrelevant information.Methods To this end,we propose MKEAH:Multimodal Knowledge Extraction and Accumulation on Hyperplanes.To ensure that the lengths of the feature vectors projected onto the hyperplane compare equally and to filter out sufficient topic-irrelevant information,two losses are proposed to learn the triplet representations from the complementary views:range loss and orthogonal loss.To interpret the capability of extracting topic-related knowledge,we present the Topic Similarity(TS)between topic and entity-relations.Results Experimental results demonstrate the effectiveness of hyperplane embedding for knowledge representation in knowledge-based visual question answering.Our model outperformed state-of-the-art methods by 2.12%and 3.24%on two challenging knowledge-request datasets:OK-VQA and KRVQA,respectively.Conclusions The obvious advantages of our model in TS show that using hyperplane embedding to represent multimodal knowledge can improve its ability to extract topic-related knowledge.
基金supported by the National Natural Science Foundation of China(Nos.62572017,62441232,62206007)R&D Program of Beijing Municipal Education Commission(KZ202210005008).
文摘Knowledge-based VisualQuestion Answering(VQA)requires the integration of visual information with external knowledge reasoning.Existing approaches typically retrieve information from external corpora and rely on pretrained language models for reasoning.However,their performance is often hindered by the limited capabilities of retrievers and the constrained size of knowledge bases.Moreover,relying on image captions to bridge the modal gap between visual and language modalities can lead to the omission of critical visual details.To address these limitations,we propose the Reflective Chain-of-Thought(ReCoT)method,a simple yet effective framework inspired by metacognition theory.ReCoT effectively activates the reasoning capabilities ofMultimodal Large LanguageModels(MLLMs),providing essential visual and knowledge cues required to solve complex visual questions.It simulates a metacognitive reasoning process that encompasses monitoring,reflection,and correction.Specifically,in the initial generation stage,an MLLM produces a preliminary answer that serves as the model’s initial cognitive output.During the reflective reasoning stage,this answer is critically examined to generate a reflective rationale that integrates key visual evidence and relevant knowledge.In the final refinement stage,a smaller language model leverages this rationale to revise the initial prediction,resulting in amore accurate final answer.By harnessing the strengths ofMLLMs in visual and knowledge grounding,ReCoT enables smaller language models to reason effectively without dependence on image captions or external knowledge bases.Experimental results demonstrate that ReCoT achieves substantial performance improvements,outperforming state-of-the-art methods by 2.26%on OK-VQA and 5.8%on A-OKVQA.
基金Supported by the Program for Liaoning Excellent Talents in University(No.LR15045)the Liaoning Provincial Science and Technology Department Applied Basic Research Plan(No.101300243).
文摘Medical visual question answering(MedVQA)faces unique challenges due to the high precision required for images and the specialized nature of the questions.These challenges include insufficient feature extraction capabilities,a lack of textual priors,and incomplete information fusion and interaction.This paper proposes an enhanced bootstrapping language-image pre-training(BLIP)model for MedVQA based on multimodal feature augmentation and triple-path collaborative attention(FCA-BLIP)to address these issues.First,FCA-BLIP employs a unified bootstrap multimodal model architecture that integrates ResNet and bidirectional encoder representations from Transformer(BERT)models to enhance feature extraction capabilities.It enables a more precise analysis of the details in images and questions.Next,the pre-trained BLIP model is used to extract features from image-text sample pairs.The model can understand the semantic relationships and shared information between images and text.Finally,a novel attention structure is developed to fuse the multimodal feature vectors,thereby improving the alignment accuracy between modalities.Experimental results demonstrate that the proposed method performs well in clinical visual question-answering tasks.For the MedVQA task of staging diabetic macular edema in fundus imaging,the proposed method outperforms the existing major models in several performance metrics.
文摘Aiming at the lack of professional knowledge to guide apparel recommendation,an apparel recommendation method based on image design expert knowledge has been proposed.Then,apparel recommendation knowledge graphs have been created and a apparel recommendation question and answer(Q&A)system has been designed and implemented.The question templates in the apparel recommendation domain were defined,the task of recognizing the named entities of question sentences was completed by the Bi-directional encoder representations from transformer-Bi-directional long short-term memory-conditional random field(BERT-BiLSTM-CRF)model,and the question template with the highest matching degree to the user’s question was obtained by using term frequency-inverse document frequency(TF-IDF)algorithm.The corresponding cypher graph database query statement was generated to retrieve the knowledge graph for answers,and iFLYTEK’s voice application programming interface(API)was called to implement the Q&A.The experimental results have shown that the Q&A system has a high accuracy rate and application value in the field of apparel recommendations.
基金Microsoft Research Asia Internet Services in Academic Research Fund(No.FY07-RES-OPP-116)the Science and Technology Development Program of Tianjin(No.06YFGZGX05900)
文摘To improve question answering （QA） performance based on real-world web data sets,a new set of question classes and a general answer re-ranking model are defined.With pre-defined dictionary and grammatical analysis,the question classifier draws both semantic and grammatical information into information retrieval and machine learning methods in the form of various training features,including the question word,the main verb of the question,the dependency structure,the position of the main auxiliary verb,the main noun of the question,the top hypernym of the main noun,etc.Then the QA query results are re-ranked by question class information.Experiments show that the questions in real-world web data sets can be accurately classified by the classifier,and the QA results after re-ranking can be obviously improved.It is proved that with both semantic and grammatical information,applications such as QA, built upon real-world web data sets, can be improved,thus showing better performance.
文摘Knowledge-based visual question answering(KB-VQA),requiring external world knowledge beyond the image for reasoning,is more challenging than traditional visual question answering.Recent works have demonstrated the effectiveness of using a large(vision)language model as an implicit knowledge source to acquire the necessary information.However,the knowledge stored in large models(LMs)is often coarse-grained and inaccurate,causing questions requiring finer-grained information to be answered incorrectly.In this work,we propose a variational expectation-maximization(EM)framework that bootstraps the VQA performance of LMs with its own answer.In contrast to former VQA pipelines,we treat the outside knowledge as a latent variable.In the E-step,we approximate the posterior with two components:First,a rough answer,e.g.,a general description of the image,which is usually the strength of LMs,and second,a multi-modal neural retriever to retrieve question-specific knowledge from an external knowledge base.In the M-step,the training objective optimizes the ability of the original LMs to generate rough answers as well as refined answers based on the retrieved information.Extensive experiments show that our proposed framework,BootLM,has a strong retrieval ability and achieves state-of-the-art performance on knowledge-based VQA tasks.
文摘Based on personal observation and students’ self-reports, this article studies the characteristics presented by students in answering questions. It is found that Chinese culture plays an important part in explaining their specific behaviors. Hence, in order to lessen the negative effect of it, the author offers her own suggestions.
基金Supported by the National Basic Research Program of China (2003CB317002)the Grant from City University of Hong Kong (7002137)
文摘A passage retrieval strategy for web-based question answering （QA） systems is proposed in our QA system. It firstly analyzes the question based on semantic patterns to obtain its syntactic and semantic information and then form initial queries. The queries are used to retrieve documents from the World Wide Web （WWW） using the Google search engine. The queries are then rewritten to form queries for passage retrieval in order to improve the precision. The relations between keywords in the question are employed in our query rewrite method. The experimental result on the question set of the TREC-2003 passage task shows that our system performs well for factoid questions.
文摘Over the last couple of decades,community question-answering sites(CQAs)have been a topic of much academic interest.Scholars have often leveraged traditional machine learning(ML)and deep learning(DL)to explore the ever-growing volume of content that CQAs engender.To clarify the current state of the CQA literature that has used ML and DL,this paper reports a systematic literature review.The goal is to summarise and synthesise the major themes of CQA research related to(i)questions,(ii)answers and(iii)users.The final review included 133 articles.Dominant research themes include question quality,answer quality,and expert identification.In terms of dataset,some of the most widely studied platforms include Yahoo!Answers,Stack Exchange and Stack Overflow.The scope of most articles was confined to just one platform with few cross-platform investigations.Articles with ML outnumber those with DL.Nonetheless,the use of DL in CQA research is on an upward trajectory.A number of research directions are proposed.
基金This work was supported in part by the National Natural Science Foundation of China under Grant 61572326,and Grant 61802258the Natural Science Foundation of Shanghai under Grant 18ZR1428300the Shanghai Committee of Science and Technology under Grant 17070502800 and Grant 16JC1403000.
文摘Deep learning models have been shown to have great advantages in answer selection tasks.The existing models,which employ encoder-decoder recurrent neural network(RNN),have been demonstrated to be effective.However,the traditional RNN-based models still suffer from limitations such as 1)high-dimensional data representation in natural language processing and 2)biased attentive weights for subsequent words in traditional time series models.In this study,a new answer selection model is proposed based on the Bidirectional Long Short-Term Memory(Bi-LSTM)and attention mechanism.The proposed model is able to generate the more effective question-answer pair representation.Experiments on a question answering dataset that includes information from multiple fields show the great advantages of our proposed model.Specifically,we achieve a maximum improvement of 3.8%over the classical LSTM model in terms of mean average precision.
基金the Specialized Research Program Fundthe Doctoral Program of Higher Education of China (20050007023)the Natural Science Foundation of Shandong Province(Y2004G04)
文摘In Chinese question answering system, because there is more semantic relation in questions than that in query words, the precision can be improved by expanding query while using natural language questions to retrieve documents. This paper proposes a new approach to query expansion based on semantics and statistics Firstly automatic relevance feedback method is used to generate a candidate expansion word set. Then the expanded query words are selected from the set based on the semantic similarity and seman- tic relevancy between the candidate words and the original words. Experiments show the new approach is effective for Web retrieval and out-performs the conventional expansion approaches.
基金the National Natural Science Foundation of China(No.61303094)the Program of Science and Technology Commission of Shanghai Municipality(Nos.16511102400 and 16111107801)the Innovation Program of Shanghai Municipal Education Commission(No.14YZ024)
文摘Question answering systems offer a friendly interface for human beings to interact with massive online information. It is time consuming for users to retrieve useful medical information with search engines among massive online websites. An effort is made to build a Chinese Question Answering System in Medical Domain(CQASMD) to provide useful medical information for users. A large medical knowledge base with more than 300 thousand medical terms and their descriptions is firstly constructed to store the structured medical knowledge data, and classified with the FastText model. Furthermore, a Word2Vec model is adopted to capture the semantic meanings of words, and the questions and answers are processed with sentence embedding to capture semantic context information. Users' questions are firstly classified and processed into a sentence vector and a matching algorithm is adopted to match the most similar question. After querying the constructed medical knowledge base, the corresponding answers to previous questions are responded to users. The architecture and flowchart of CQASMD is proposed, which will play an important role in self disease diagnosis and treatment.
基金The National Natural Science Foundation of China(No.61502095).
文摘Aiming at the relation linking task for question answering over knowledge base,especially the multi relation linking task for complex questions,a relation linking approach based on the multi-attention recurrent neural network(RNN)model is proposed,which works for both simple and complex questions.First,the vector representations of questions are learned by the bidirectional long short-term memory(Bi-LSTM)model at the word and character levels,and named entities in questions are labeled by the conditional random field(CRF)model.Candidate entities are generated based on a dictionary,the disambiguation of candidate entities is realized based on predefined rules,and named entities mentioned in questions are linked to entities in knowledge base.Next,questions are classified into simple or complex questions by the machine learning method.Starting from the identified entities,for simple questions,one-hop relations are collected in the knowledge base as candidate relations;for complex questions,two-hop relations are collected as candidates.Finally,the multi-attention Bi-LSTM model is used to encode questions and candidate relations,compare their similarity,and return the candidate relation with the highest similarity as the result of relation linking.It is worth noting that the Bi-LSTM model with one attentions is adopted for simple questions,and the Bi-LSTM model with two attentions is adopted for complex questions.The experimental results show that,based on the effective entity linking method,the Bi-LSTM model with the attention mechanism improves the relation linking effectiveness of both simple and complex questions,which outperforms the existing relation linking methods based on graph algorithm or linguistics understanding.
文摘ExpertRecommendation(ER)aims to identify domain experts with high expertise and willingness to provide answers to questions in Community Question Answering(CQA)web services.How to model questions and users in the heterogeneous content network is critical to this task.Most traditional methods focus on modeling questions and users based on the textual content left in the community while ignoring the structural properties of heterogeneous CQA networks and always suffering from textual data sparsity issues.Recent approaches take advantage of structural proximities between nodes and attempt to fuse the textual content of nodes for modeling.However,they often fail to distinguish the nodes’personalized preferences and only consider the textual content of a part of the nodes in network embedding learning,while ignoring the semantic relevance of nodes.In this paper,we propose a novel framework that jointly considers the structural proximity relations and textual semantic relevance to model users and questions more comprehensively.Specifically,we learn topology-based embeddings through a hierarchical attentive network learning strategy,in which the proximity information and the personalized preference of nodes are encoded and preserved.Meanwhile,we utilize the node’s textual content and the text correlation between adjacent nodes to build the content-based embedding through a meta-context-aware skip-gram model.In addition,the user’s relative answer quality is incorporated to promote the ranking performance.Experimental results show that our proposed framework consistently and significantly outperforms the state-of-the-art baselines on three real-world datasets by taking the deep semantic understanding and structural feature learning together.The performance of the proposed work is analyzed in terms of MRR,P@K,and MAP and is proven to be more advanced than the existing methodologies.
基金Project(61702063)supported by the National Natural Science Foundation of China。
文摘With the warming up and continuous development of machine learning,especially deep learning,the research on visual question answering field has made significant progress,with important theoretical research significance and practical application value.Therefore,it is necessary to summarize the current research and provide some reference for researchers in this field.This article conducted a detailed and in-depth analysis and summarized of relevant research and typical methods of visual question answering field.First,relevant background knowledge about VQA(Visual Question Answering)was introduced.Secondly,the issues and challenges of visual question answering were discussed,and at the same time,some promising discussion on the particular methodologies was given.Thirdly,the key sub-problems affecting visual question answering were summarized and analyzed.Then,the current commonly used data sets and evaluation indicators were summarized.Next,in view of the popular algorithms and models in VQA research,comparison of the algorithms and models was summarized and listed.Finally,the future development trend and conclusion of visual question answering were prospected.
文摘In this work, a best answer recommendation model is proposed for a Question Answering (QA) system. A Community Question Answering System was subsequently developed based on the model. The system applies Brouwer Fixed Point Theorem to prove the existence of the desired voter scoring function and Normalized Google Distance (NGD) to show closeness between words before an answer is suggested to users. Answers are ranked according to their Fixed-Point Score (FPS) for each question. Thereafter, the highest scored answer is chosen as the FPS Best Answer (BA). For each question asked by user, the system applies NGD to check if similar or related questions with the best answer had been asked and stored in the database. When similar or related questions with the best answer are not found in the database, Brouwer Fixed point is used to calculate the best answer from the pool of answers on a question then the best answer is stored in the NGD data-table for recommendation purpose. The system was implemented using PHP scripting language, MySQL for database management, JQuery, and Apache. The system was evaluated using standard metrics: Reciprocal Rank, Mean Reciprocal Rank (MRR) and Discounted Cumulative Gain (DCG). The system eliminated longer waiting time faced by askers in a community question answering system. The developed system can be used for research and learning purposes.
基金supported by the National Research Foundation of Korea(NRF)grant funded by the Korea Government(MSIT)(No.2020R1G1A1100493).
文摘Recently,pre-trained language representation models such as bidirec-tional encoder representations from transformers(BERT)have been performing well in commonsense question answering(CSQA).However,there is a problem that the models do not directly use explicit information of knowledge sources existing outside.To augment this,additional methods such as knowledge-aware graph network(KagNet)and multi-hop graph relation network(MHGRN)have been proposed.In this study,we propose to use the latest pre-trained language model a lite bidirectional encoder representations from transformers(ALBERT)with knowledge graph information extraction technique.We also propose to applying the novel method,schema graph expansion to recent language models.Then,we analyze the effect of applying knowledge graph-based knowledge extraction techniques to recent pre-trained language models and confirm that schema graph expansion is effective in some extent.Furthermore,we show that our proposed model can achieve better performance than existing KagNet and MHGRN models in CommonsenseQA dataset.
文摘The original intention of visual question answering(VQA)models is to infer the answer based on the relevant information of the question text in the visual image,but many VQA models often yield answers that are biased by some prior knowledge,especially the language priors.This paper proposes a mitigation model called language priors mitigation-VQA(LPM-VQA)for the language priors problem in VQA model,which divides language priors into positive and negative language priors.Different network branches are used to capture and process the different priors to achieve the purpose of mitigating language priors.A dynamically-changing language prior feedback objective function is designed with the intermediate results of some modules in the VQA model.The weight of the loss value for each answer is dynamically set according to the strength of its language priors to balance its proportion in the total VQA loss to further mitigate the language priors.This model does not depend on the baseline VQA architectures and can be configured like a plug-in to improve the performance of the model over most existing VQA models.The experimental results show that the proposed model is general and effective,achieving state-of-the-art accuracy in the VQA-CP v2 dataset.
文摘In the context of power generation companies, vast amounts of specialized data and expert knowledge have been accumulated. However, challenges such as data silos and fragmented knowledge hinder the effective utilization of this information. This study proposes a novel framework for intelligent Question-and-Answer (Q&A) systems based on Retrieval-Augmented Generation (RAG) to address these issues. The system efficiently acquires domain-specific knowledge by leveraging external databases, including Relational Databases (RDBs) and graph databases, without additional fine-tuning for Large Language Models (LLMs). Crucially, the framework integrates a Dynamic Knowledge Base Updating Mechanism (DKBUM) and a Weighted Context-Aware Similarity (WCAS) method to enhance retrieval accuracy and mitigate inherent limitations of LLMs, such as hallucinations and lack of specialization. Additionally, the proposed DKBUM dynamically adjusts knowledge weights within the database, ensuring that the most recent and relevant information is utilized, while WCAS refines the alignment between queries and knowledge items by enhanced context understanding. Experimental validation demonstrates that the system can generate timely, accurate, and context-sensitive responses, making it a robust solution for managing complex business logic in specialized industries.
