Browse
Submit Data
Databases
API
Help

Dataset Information

11 Views

0 Connections

0 Citations

0 Reanalyses

0 Downloads

Omics score: 0

Multi-Hop Question Generation Using Hierarchical Encoding-Decoding and Context Switch Mechanism.

ABSTRACT: Neural auto-regressive sequence-to-sequence models have been dominant in text generation tasks, especially the question generation task. However, neural generation models suffer from the global and local semantic semantic drift problems. Hence, we propose the hierarchical encoding-decoding mechanism that aims at encoding rich structure information of the input passages and reducing the variance in the decoding phase. In the encoder, we hierarchically encode the input passages according to its structure at four granularity-levels: [word, chunk, sentence, document]-level. Second, we progressively select the context vector from the document-level representations to the word-level representations at each decoding time step. At each time-step in the decoding phase, we progressively select the context vector from the document-level representations to word-level. We also propose the context switch mechanism that enables the decoder to use the context vector from the last step when generating the current word at each time-step.It provides a means of improving the stability of the text generation process during the decoding phase when generating a set of consecutive words. Additionally, we inject syntactic parsing knowledge to enrich the word representations. Experimental results show that our proposed model substantially improves the performance and outperforms previous baselines according to both automatic and human evaluation. Besides, we implement a deep and comprehensive analysis of generated questions based on their types.

SUBMITTER: Ji T

PROVIDER: S-EPMC8618393 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Json Xml

Similar Datasets

Multi-modal adaptive gated mechanism for visual question answering.

Project description:Visual Question Answering (VQA) is a multimodal task that uses natural language to ask and answer questions based on image content. For multimodal tasks, obtaining accurate modality feature information is crucial. The existing researches on the visual question answering model mainly start from the perspective of attention mechanism and multimodal fusion, which will tend to ignore the impact of modal interaction learning and the introduction of noise information in the process of modal fusion on the overall performance of the model. This paper proposes a novel and efficient multimodal adaptive gated mechanism model, MAGM. The model adds an adaptive gate mechanism to the intra- and inter-modality learning and the modal fusion process. This model can effectively filter irrelevant noise information, obtain fine-grained modal features, and improve the ability of the model to adaptively control the contribution of the two modal features to the predicted answer. In intra- and inter-modality learning modules, the self-attention gated and self-guided-attention gated units are designed to filter text and image features' noise information effectively. In modal fusion module, the adaptive gated modal feature fusion structure is designed to obtain fine-grained modal features and improve the accuracy of the model in answering questions. Quantitative and qualitative experiments on the two VQA task benchmark datasets, VQA 2.0 and GQA, proved that the method in this paper is superior to the existing methods. The MAGM model has an overall accuracy of 71.30% on the VQA 2.0 dataset and an overall accuracy of 57.57% on the GQA dataset.

| S-EPMC10306234 | biostudies-literature

Molecular encoding and synaptic decoding of context during salt chemotaxis in C. elegans.

Project description:Animals navigate toward favorable locations using various environmental cues. However, the mechanism of how the goal information is encoded and decoded to generate migration toward the appropriate direction has not been clarified. Here, we describe the mechanism of migration towards a learned concentration of NaCl in Caenorhabditis elegans. In the salt-sensing neuron ASER, the difference between the experienced and currently perceived NaCl concentration is encoded as phosphorylation at Ser65 of UNC-64/Syntaxin 1 A through the protein kinase C(PKC-1) signaling pathway. The phosphorylation affects basal glutamate transmission from ASER, inducing the reversal of the postsynaptic response of reorientation-initiating neurons (i.e., from inhibitory to excitatory), guiding the animals toward the experienced concentration. This process, the decoding of the context, is achieved through the differential sensitivity of postsynaptic excitatory and inhibitory receptors. Our results reveal the mechanism of migration based on the synaptic plasticity that conceptually differs from the classical ones.

| S-EPMC9142520 | biostudies-literature

Hierarchical Encoding of Attended Auditory Objects in Multi-talker Speech Perception.

Project description:Humans can easily focus on one speaker in a multi-talker acoustic environment, but how different areas of the human auditory cortex (AC) represent the acoustic components of mixed speech is unknown. We obtained invasive recordings from the primary and nonprimary AC in neurosurgical patients as they listened to multi-talker speech. We found that neural sites in the primary AC responded to individual speakers in the mixture and were relatively unchanged by attention. In contrast, neural sites in the nonprimary AC were less discerning of individual speakers but selectively represented the attended speaker. Moreover, the encoding of the attended speaker in the nonprimary AC was invariant to the degree of acoustic overlap with the unattended speaker. Finally, this emergent representation of attended speech in the nonprimary AC was linearly predictable from the primary AC responses. Our results reveal the neural computations underlying the hierarchical formation of auditory objects in human AC during multi-talker speech perception.

| S-EPMC8082956 | biostudies-literature

An Efficient Next Hop Selection Algorithm for Multi-Hop Body Area Networks.

Project description:Body Area Networks (BANs) consist of various sensors which gather patient's vital signs and deliver them to doctors. One of the most significant challenges faced, is the design of an energy-efficient next hop selection algorithm to satisfy Quality of Service (QoS) requirements for different healthcare applications. In this paper, a novel efficient next hop selection algorithm is proposed in multi-hop BANs. This algorithm uses the minimum hop count and a link cost function jointly in each node to choose the best next hop node. The link cost function includes the residual energy, free buffer size, and the link reliability of the neighboring nodes, which is used to balance the energy consumption and to satisfy QoS requirements in terms of end to end delay and reliability. Extensive simulation experiments were performed to evaluate the efficiency of the proposed algorithm using the NS-2 simulator. Simulation results show that our proposed algorithm provides significant improvement in terms of energy consumption, number of packets forwarded, end to end delay and packet delivery ratio compared to the existing routing protocol.

| S-EPMC4714909 | biostudies-literature

Rats answer an unexpected question after incidental encoding.

Project description:A fundamental aspect of episodic memory is that retrieval of information can occur when encoding is incidental and memory assessment is unexpected. These features are difficult to model in animals because behavioral training likely gives rise to well-learned expectations about the sequence of events. Thus, the possibility remains that animals may solve an episodic memory test by using well-learned semantic rules without remembering the episode at memory assessment. Here we show that rats can answer an unexpected question after incidental encoding in a hippocampal-dependent manner, consistent with the use of episodic memory. Rats were initially trained to report about a recent event (food versus no food) and separately searched for food where there was no expectation of being asked about the presence of food. To test episodic memory, we gave rats the opportunity to incidentally encode the presence or absence of food and unexpectedly asked them to report about the recent event. Temporary inactivation of the CA3 region of the hippocampus with bilateral infusions of lidocaine selectively eliminated the ability of rats to answer the unexpected, but not the expected, question. Our studies suggest that rats remember an earlier episode after incidental encoding based upon hippocampal-dependent episodic memory.

| S-EPMC3376895 | biostudies-literature

Real-time decoding of question-and-answer speech dialogue using human cortical activity.

Project description:Natural communication often occurs in dialogue, differentially engaging auditory and sensorimotor brain regions during listening and speaking. However, previous attempts to decode speech directly from the human brain typically consider listening or speaking tasks in isolation. Here, human participants listened to questions and responded aloud with answers while we used high-density electrocorticography (ECoG) recordings to detect when they heard or said an utterance and to then decode the utterance's identity. Because certain answers were only plausible responses to certain questions, we could dynamically update the prior probabilities of each answer using the decoded question likelihoods as context. We decode produced and perceived utterances with accuracy rates as high as 61% and 76%, respectively (chance is 7% and 20%). Contextual integration of decoded question likelihoods significantly improves answer decoding. These results demonstrate real-time decoding of speech in an interactive, conversational setting, which has important implications for patients who are unable to communicate.

| S-EPMC6667454 | biostudies-literature

Decoding hierarchical control of sequential behavior in oscillatory EEG activity.

Project description:Despite strong theoretical reasons for assuming that abstract representations organize complex action sequences in terms of subplans (chunks) and sequential positions, we lack methods to directly track such content-independent, hierarchical representations in humans. We applied time-resolved, multivariate decoding analysis to the pattern of rhythmic EEG activity that was registered while participants planned and executed individual elements from pre-learned, structured sequences. Across three experiments, the theta and alpha-band activity coded basic elements and abstract control representations, in particular, the ordinal position of basic elements, but also the identity and position of chunks. Further, a robust representation of higher level, chunk identity information was only found in individuals with above-median working memory capacity, potentially providing a neural-level explanation for working-memory differences in sequential performance. Our results suggest that by decoding oscillatory activity we can track how the cognitive system traverses through the states of a hierarchical control structure.

| S-EPMC6257819 | biostudies-literature

Structured neuronal encoding and decoding of human speech features.

Project description:Human speech sounds are produced through a coordinated movement of structures along the vocal tract. Here we show highly structured neuronal encoding of vowel articulation. In medial-frontal neurons, we observe highly specific tuning to individual vowels, whereas superior temporal gyrus neurons have nonspecific, sinusoidally modulated tuning (analogous to motor cortical directional tuning). At the neuronal population level, a decoding analysis reveals that the underlying structure of vowel encoding reflects the anatomical basis of articulatory movements. This structured encoding enables accurate decoding of volitional speech segments and could be applied in the development of brain-machine interfaces for restoring speech in paralysed individuals.

| S-EPMC4445934 | biostudies-literature

Decoding selective attention to context memory: An aging study.

Project description:Emerging evidence has suggested that the tendency for older adults to bind too much contextual information during encoding (i.e., hyper-binding) may contribute to poorer memory for relevant contextual information during retrieval. While these findings are consistent with theories of age-related declines in selective attention and inhibitory control, the degree to which older adults are able to selectively attend to relevant contextual information during encoding is unknown. To better understand the neural dynamics associated with selective attention during encoding, the current study applied multivariate pattern analyses (MVPA) to oscillatory EEG in order to track moment-to-moment shifts of attention between relevant and irrelevant contextual information during encoding. Young and older adults studied pictures of objects in the presence of two contextual features: a color and a scene, and their attention was directed to the object's relationship with one of those contexts (i.e., target context). Results showed that patterns of oscillatory power successfully predicted whether selective attention was directed to a scene or color, across age groups. Individual differences in overall classification performance were associated with individual differences in target context memory accuracy during retrieval. However, changes in classification performance within a trial, suggestive of fluctuations in selective attention, predicted individual differences in hyper-binding. To the best of our knowledge, this is the first study to use MPVA techniques to decode attention during episodic encoding and the impact of attentional shifts toward distracting information on age-related context memory impairments and hyper-binding. These results are consistent with the as-of-yet unsubstantiated theory that age-related declines in context memory may be attributable to poorer selective attention and/or greater inhibitory deficits in older adults.

| S-EPMC6468982 | biostudies-literature

Decoding ALS: from genes to mechanism.

Project description:Amyotrophic lateral sclerosis (ALS) is a progressive and uniformly fatal neurodegenerative disease. A plethora of genetic factors have been identified that drive the degeneration of motor neurons in ALS, increase susceptibility to the disease or influence the rate of its progression. Emerging themes include dysfunction in RNA metabolism and protein homeostasis, with specific defects in nucleocytoplasmic trafficking, the induction of stress at the endoplasmic reticulum and impaired dynamics of ribonucleoprotein bodies such as RNA granules that assemble through liquid-liquid phase separation. Extraordinary progress in understanding the biology of ALS provides new reasons for optimism that meaningful therapies will be identified.

| S-EPMC5585017 | biostudies-literature

OmicsDI is part of the ELIXIR infrastructure

OmicsDI is an Elixir interoperability service. Learn more ›

Tweets

OmicsDI Databases

PRIDE
PeptideAtlas
MassIVE
JPOST Repository
Physiome Model Repository

EGA
EVA
ENA
LINCS
PAXDB
Cell Collective

MetaboLights
Metabolomics Workbench
MetabolomeExpress
GNPS
BioModels
FAIRDOMHub

ArrayExpress
dbGaP
ExpressionAtlas
GEO
NODE

Information

Databases
Help
API
Contact us
Code on GitHub
Terms of use
Submit Data