20,976 Hits in 8.9 sec

A Thousand Words Are Worth More Than a Picture: Natural Language-Centric Outside-Knowledge Visual Question Answering [article]

Feng Gao, Qing Ping, Govind Thattai, Aishwarya Reganti, Ying Nian Wu, Prem Natarajan
2022 arXiv   pre-print
Outside-knowledge visual question answering (OK-VQA) requires the agent to comprehend the image, make use of relevant knowledge from the entire web, and digest all the information to answer the question  ...  In this paper, we call for a paradigm shift for the OK-VQA task, which transforms the image into plain text, so that we can enable knowledge passage retrieval, and generative question-answering in the  ...  Outside Knowledge-Based VQA (OK-VQA) Beyond the above paradigm, knowledge-based visual question answering (KB-VQA) is proposed where a visual question cannot be answered without external knowledge.  ... 
arXiv:2201.05299v1 fatcat:cbr5icvimbh7xfj7m2vxmpjyma

Modern Question Answering Datasets and Benchmarks: A Survey [article]

Zhen Wang
2022 arXiv   pre-print
Specifically, we begin with introducing two of the most common QA tasks - textual question answer and visual question answering - separately, covering the most representative datasets, and then give some  ...  Question Answering (QA) is one of the most important natural language processing (NLP) tasks.  ...  For each sample in OK-VQA (Marino et al., 2019) , in addition to a question and an answer, they also provide some outside knowledge, in which the correct answer contained.  ... 
arXiv:2206.15030v1 fatcat:zsrty2ifxzhcfoplvn6vtebeui

Can Open Domain Question Answering Systems Answer Visual Knowledge Questions? [article]

Jiawen Zhang, Abhijit Mishra, Avinesh P.V.S, Siddharth Patwardhan, Sachin Agarwal
2022 arXiv   pre-print
The task of Outside Knowledge Visual Question Answering (OKVQA) requires an automatic system to answer natural language questions about pictures and images using external knowledge.  ...  This allows for the reuse of existing text-based Open Domain Question Answering (QA) Systems for visual question answering.  ...  ., Wikipedia) or knowledge representations from massive KGs are first retrieved and then comprehended by machinelearned models to yield potential answers from the extracted passages.  ... 
arXiv:2202.04306v1 fatcat:ycnfaljzkfcltggmwrzcbjdite

Select, Substitute, Search: A New Benchmark for Knowledge-Augmented Visual Question Answering [article]

Aman Jain, Mayank Kothyari, Vishwajeet Kumar, Preethi Jyothi, Ganesh Ramakrishnan, Soumen Chakrabarti
2021 arXiv   pre-print
Multimodal IR, spanning text corpus, knowledge graph and images, called outside knowledge visual question answering (OKVQA), is of much recent interest.  ...  Specifically, the questioner identifies an entity in the image and asks a question involving that entity which can be answered only by consulting a knowledge graph or corpus passage mentioning the entity  ...  VQA systems [25] have started incorporating external knowledge for question answering.  ... 
arXiv:2103.05568v2 fatcat:qgtniebl3fcthicawi5pf6jccm

LaKo: Knowledge-driven Visual Question Answering via Late Knowledge-to-Text Injection [article]

Zhuo Chen, Yufeng Huang, Jiaoyan Chen, Yuxia Geng, Yin Fang, Jeff Pan, Ningyu Zhang, Wen Zhang
2022 arXiv   pre-print
Visual question answering (VQA) often requires an understanding of visual concepts and language semantics, which relies on external knowledge.  ...  In this paper, we propose LaKo, a knowledge-driven VQA method via Late Knowledge-to-text Injection.  ...  Knowledge-based VQA The knowledge-based VQA requires the model to acquire factual or common-sense knowledge outside the image for question answering.  ... 
arXiv:2207.12888v1 fatcat:ru2vqlszxvb65laeokebt76itu

A Survey of Text Mining Techniques and Applications

Vishal Gupta, Gurpreet S. Lehal
2009 Journal of Emerging Technologies in Web Intelligence  
Next, the Answer Extraction obtains the answer from the expected type, constraints and passages returned by Question Analysis and Passage Retrieval modules. The architecture is shown in figure15.  ...  In Passage Retrieval module, user question is handed over also to the JIRS Passage Retrieval system, more specifically to its Search Engine and N-grams Extraction components.  ...  He has developed a number of research projects in field of NLP including synonyms detection, automatic question answering and text summarization etc.  ... 
doi:10.4304/jetwi.1.1.60-76 fatcat:wx3k6xjfh5dodjewrpu2jgfs4y

Studying the Effect of Retrieval Direction during Reading on Productive and Receptive Knowledge of Vocabulary

Sasan Baleghizadeh, Samaneh Shafeie
2017 Issues in language teaching  
A Posttest was conducted in both productive and receptive directions for all the groups.  ...  Retrieval tasks provide learners with an opportunity to focus both on meaning and on form. There are four different retrieval directions.  ...  Reading passages were gathered and participants were given a sheet of math problems to answer as many as they can in five minutes for distracting purposes.  ... 
doi:10.22054/ilt.2019.23022.206 doaj:e72d7dcd5ba74522b0d5ab13c480fc80 fatcat:v7wzpky2kvbd5jizzhnqd5kaq4

A Wrong Answer or a Wrong Question? An Intricate Relationship between Question Reformulation and Answer Selection in Conversational Question Answering [article]

Svitlana Vakulenko, Shayne Longpre, Zhucheng Tu, Raviteja Anantha
2022 arXiv   pre-print
Our experiments uncover sensitivity to question formulation of the popular state-of-the-art models for reading comprehension and passage ranking.  ...  Our results demonstrate that the reading comprehension model is insensitive to question formulation, while the passage ranking changes dramatically with a little variation in the input question.  ...  Acknowledgements We would like to thank our colleagues Srinivas Chappidi, Bjorn Hoffmeister, Stephan Peitz, Russ Webb, Drew Frank, and Chris DuBois for their insightful comments.  ... 
arXiv:2010.06835v2 fatcat:42e77ad2a5ckhozqy77mcyhddi

MKQA: A Linguistically Diverse Benchmark for Multilingual Open Domain Question Answering [article]

Shayne Longpre, Yi Lu, Joachim Daiber
2021 arXiv   pre-print
We introduce Multilingual Knowledge Questions and Answers (MKQA), an open-domain question answering evaluation set comprising 10k question-answer pairs aligned across 26 typologically diverse languages  ...  With 26 languages, this dataset supplies the widest range of languages to-date for evaluating question answering.  ...  Mendes and Charles Srisuwananukorn for guidance and support, as well as to Noriyo Sakamoto for help in data collection.  ... 
arXiv:2007.15207v2 fatcat:pmnbniyko5cjzbgv5o4hvjhpta

MKQA: A Linguistically Diverse Benchmark for Multilingual Open Domain Question Answering

Shayne Longpre, Yi Lu, Joachim Daiber
2021 Transactions of the Association for Computational Linguistics  
We introduce Multilingual Knowledge Questions and Answers (MKQA), an open- domain question answering evaluation set comprising 10k question-answer pairs aligned across 26 typologically diverse languages  ...  With 26 languages, this dataset supplies the widest range of languages to-date for evaluating question answering.  ...  Mendes and Charles Srisuwananukorn for guidance and support, as well as to Noriyo Sakamoto for help in data collection.  ... 
doi:10.1162/tacl_a_00433 fatcat:hef2fubo7jfjjftxt6juo5uuf4

The pretesting effect: Do unsuccessful retrieval attempts enhance learning?

Lindsey E. Richland, Nate Kornell, Liche Sean Kao
2009 Journal of experimental psychology. Applied  
but not asking participants to answer them before reading the passage.  ...  In the test condition, they were asked about embedded concepts before reading the passage; in the extended study condition, they were given a longer time to read the passage.  ...  To retrieve an appropriate answer to a question, a learner may attempt to imagine or creatively search for potential solutions.  ... 
doi:10.1037/a0016496 pmid:19751074 fatcat:j4hsqwx7fvcmjjpsiwgj52mwnu

WebQA: Multihop and Multimodal QA [article]

Yingshan Chang, Mridu Narang, Hisami Suzuki, Guihong Cao, Jianfeng Gao, Yonatan Bisk
2022 arXiv   pre-print
Scaling Visual Question Answering (VQA) to the open-domain and multi-hop nature of web searches, requires fundamental advances in visual representation learning, knowledge aggregation, and language generation  ...  knowledge, but also the richer visual online world.  ...  for image-based question answering.  ... 
arXiv:2109.00590v4 fatcat:mqw3jyj3hfasbkhxghwp6vpoce

Artificial Intelligence (AI) in Action: Addressing the COVID-19 Pandemic with Natural Language Processing (NLP) [article]

Qingyu Chen, Robert Leaman, Alexis Allot, Ling Luo, Chih-Hsuan Wei, Shankai Yan, Zhiyong Lu
2021 arXiv   pre-print
We detail work on four core NLP tasks: information retrieval, named entity recognition, literature-based discovery, and question answering.  ...  Many of these difficulties are fundamentally information needs; attempts to address these needs have caused an information overload for both researchers and the public.  ...  Document/passage retrieval This module retrieves the documents/passages that are relevant to the processed questions.  ... 
arXiv:2010.16413v2 fatcat:cxzxmnnbfrednpto4zkpnemltm

Profiling classroom reading comprehension development practices from the PIRLS 2006 in South Africa

Lisa Zimmerman, Brigitte Smit
2014 South African Journal of Education  
To shed light on potential reasons for learners' reading comprehension difficulties, this article presents selected findings on teachers' reading comprehension development practices emanating from the  ...  specific concern is the lack of representation of the sampled South African learners at the PIRLS international benchmarks, revealing a distinct lack of their development of thinking and reasoning abilities for  ...  retrieval questions.  ... 
doi:10.15700/201409161101 fatcat:lubqh7q4hfajhjhpuczrc3duvi

A Classroom Study on the Relationship Between Student Achievement and Retrieval-Enhanced Learning

Shana K. Carpenter, Terry J. S. Lund, Clark R. Coffman, Patrick I. Armstrong, Monica H. Lamm, Robert D. Reason
2015 Educational Psychology Review  
For example, Kang et al. (2007) found that undergraduate students' learning of scientific articles was significantly enhanced by reading the articles once and then Educ Psychol Rev completing short-answer  ...  Retrieval has been known to produce significant learning gains for a wide variety of materials, including word lists (  ...  the chances that a later question tapping knowledge of the lure would be answered correctly.  ... 
doi:10.1007/s10648-015-9311-9 fatcat:vw6q3mcavjc67fck4dgscx3t4i
« Previous Showing results 1 — 15 out of 20,976 results