Visual question answering(VQA)has attracted more and more attention in computer vision and natural language processing.Scholars are committed to studying how to better integrate image features and text features to ach...Visual question answering(VQA)has attracted more and more attention in computer vision and natural language processing.Scholars are committed to studying how to better integrate image features and text features to achieve better results in VQA tasks.Analysis of all features may cause information redundancy and heavy computational burden.Attention mechanism is a wise way to solve this problem.However,using single attention mechanism may cause incomplete concern of features.This paper improves the attention mechanism method and proposes a hybrid attention mechanism that combines the spatial attention mechanism method and the channel attention mechanism method.In the case that the attention mechanism will cause the loss of the original features,a small portion of image features were added as compensation.For the attention mechanism of text features,a selfattention mechanism was introduced,and the internal structural features of sentences were strengthened to improve the overall model.The results show that attention mechanism and feature compensation add 6.1%accuracy to multimodal low-rank bilinear pooling network.展开更多
Prostate cancer and its treatment have long-term implications for men’s lives. We aimed to describe the content, extent, and frequency of written comments to the open-ended question, “Further comments?” in the pati...Prostate cancer and its treatment have long-term implications for men’s lives. We aimed to describe the content, extent, and frequency of written comments to the open-ended question, “Further comments?” in the patient-reported outcome measures questionnaire. During the study period, 897 men participated;372 wrote 747 free-text comments in the questionnaire. These comments were analysed using qualitative content analysis and were grouped into four categories: 1) prostate cancer’s influence on health;2) clarifications of answers to the survey;3) descriptions of well-being despite the cancer;and 4) experiences of care and the need for contact with health care. The distribution of the comments shifted over time. The open-ended question not only allowed the participants to explain their other responses and describe important aspects of their lives during and after treatment, something not normally covered by a questionnaire, but it also indicated their experiences of health care services along the patients’ PC-trajectory. This further raises the issue of including an open-ended item in a forced-choice survey into the ethical realm to ensure that proper care is taken of participants’ answers and thoughts.展开更多
目前知识库问答(Knowledge base question answering,KBQA)技术无法有效地处理复杂问题,难以理解其中的复杂语义.将一个复杂问题先分解再整合,是解析复杂语义的有效方法.但是,在问题分解的过程中往往会出现实体判断错误或主题实体缺失...目前知识库问答(Knowledge base question answering,KBQA)技术无法有效地处理复杂问题,难以理解其中的复杂语义.将一个复杂问题先分解再整合,是解析复杂语义的有效方法.但是,在问题分解的过程中往往会出现实体判断错误或主题实体缺失的情况,导致分解得到的子问题与原始复杂问题并不匹配.针对上述问题,提出了一种融合事实文本的问解分解式语义解析方法.对复杂问题的处理分为分解-抽取-解析3个阶段,首先把复杂问题分解成简单子问题,然后抽取问句中的关键信息,最后生成结构化查询语句.同时,本文又构造了事实文本库,将三元组转化成用自然语言描述的句子,采用注意力机制获取更丰富的知识.在ComplexWebQuestions数据集上的实验表明,本文提出的模型在性能上优于其他基线模型.展开更多
基金This work was supported by the Sichuan Science and Technology Program(2021YFQ0003).
文摘Visual question answering(VQA)has attracted more and more attention in computer vision and natural language processing.Scholars are committed to studying how to better integrate image features and text features to achieve better results in VQA tasks.Analysis of all features may cause information redundancy and heavy computational burden.Attention mechanism is a wise way to solve this problem.However,using single attention mechanism may cause incomplete concern of features.This paper improves the attention mechanism method and proposes a hybrid attention mechanism that combines the spatial attention mechanism method and the channel attention mechanism method.In the case that the attention mechanism will cause the loss of the original features,a small portion of image features were added as compensation.For the attention mechanism of text features,a selfattention mechanism was introduced,and the internal structural features of sentences were strengthened to improve the overall model.The results show that attention mechanism and feature compensation add 6.1%accuracy to multimodal low-rank bilinear pooling network.
文摘Prostate cancer and its treatment have long-term implications for men’s lives. We aimed to describe the content, extent, and frequency of written comments to the open-ended question, “Further comments?” in the patient-reported outcome measures questionnaire. During the study period, 897 men participated;372 wrote 747 free-text comments in the questionnaire. These comments were analysed using qualitative content analysis and were grouped into four categories: 1) prostate cancer’s influence on health;2) clarifications of answers to the survey;3) descriptions of well-being despite the cancer;and 4) experiences of care and the need for contact with health care. The distribution of the comments shifted over time. The open-ended question not only allowed the participants to explain their other responses and describe important aspects of their lives during and after treatment, something not normally covered by a questionnaire, but it also indicated their experiences of health care services along the patients’ PC-trajectory. This further raises the issue of including an open-ended item in a forced-choice survey into the ethical realm to ensure that proper care is taken of participants’ answers and thoughts.
文摘目前知识库问答(Knowledge base question answering,KBQA)技术无法有效地处理复杂问题,难以理解其中的复杂语义.将一个复杂问题先分解再整合,是解析复杂语义的有效方法.但是,在问题分解的过程中往往会出现实体判断错误或主题实体缺失的情况,导致分解得到的子问题与原始复杂问题并不匹配.针对上述问题,提出了一种融合事实文本的问解分解式语义解析方法.对复杂问题的处理分为分解-抽取-解析3个阶段,首先把复杂问题分解成简单子问题,然后抽取问句中的关键信息,最后生成结构化查询语句.同时,本文又构造了事实文本库,将三元组转化成用自然语言描述的句子,采用注意力机制获取更丰富的知识.在ComplexWebQuestions数据集上的实验表明,本文提出的模型在性能上优于其他基线模型.
