Order Recall Research Articles

A preserved neural code for temporal order between memory formation and recall in the human medial temporal lobe.

Temporal memory enables us to remember the temporal order of events happening in our life. The human medial temporal lobe (MTL) appears to contain neural representations supporting temporal memory formation, but the cellular mechanisms that preserve temporal order information for recall are largely unknown. Here, we examined whether human MTL neuronal activity represents the temporal position of events during memory formation and recall, using invasive single and multi-unit recordings in human epilepsy patients (n = 19). Participants freely navigated a virtual environment in order to explore and remember locations and temporal positions of objects. During each exploration period, they sequentially encountered two or three different objects, placed in different locations. This allowed us to examine single- and multi-unit neuronal firing rates (FR) as a function of the temporal position the objects were presented in. We found that a significant number of multi-units and single-units in various MTL regions including the hippocampus showed selectivity to the temporal position of objects during the exploration period. During recall, patients were asked to indicate which one of two objects from the same trial was found latter. Neural firing rates during recall showed a selectivity supporting recall of temporal positions. Interestingly, most of the selective single-units that stayed selective during encoding and recall preserved their temporal position preference. Our results thus suggest that neuronal activity in the human MTL contains a preserved neural code for temporal order in memory formation and recall.

소비자 안전보건을 위한 ESG 공시기준에의 제품안전 지표설정 연구

ESG management has now become a new management paradigm both domestically and internationally. ESG is a non-financial performance indicator to achieve corporate sustainability, and encompasses responsibilities and obligations to stakeholders. Recently, as the safety issue of overseas products imported from online shopping malls in China has emerged, the need to ensure consumer safety and health from hazardous products is increasing. Accordingly, the establishment of product safety disclosure standards for consumer protection is required in the ESG information disclosure process. In this paper, in order to establish product safety indicators for sustainability disclosure standards for ESG management, the disclosure standards of major domestic and foreign initiatives were reviewed and analyzed to examine the introduction of product safety-related indicators, and product safety indicators that can be set for consumer safety and health were developed and presented. As product safety indicators, this paper first presented the occurrence of illegal and defective products, occurrence of product accidents, block hazardous products, number of recall orders and recommendations, and recall recovery rate as indicators that can be introduced throughout the entire product life cycle. In addition, the product risk level, the number and ratio of products for which risk assessment was performed, and the methods and standards used for risk assessment were presented as indicators related to product risk that indicate the level of impact of the product on consumer safety and health. In addition, the consumer safety index was presented as an indicator related to product safety accidents that occur during the consumer product use phase. The consumer safety index is an index that allows consumers to directly feel the level of product safety due to product accidents that occur when consumers use the product. It can be said to be a product safety index from the consumer's perspective that can protect consumer safety.

Do actions structure auditory memory? Action-based event segmentation effects on sensory responses, pupil dilation and sequential memory.

Our actions shape our everyday experience: what we experience, how we perceive, and remember it are deeply affected by how we interact with the world. Performing an action to deliver a stimulus engages neurophysiological processes which are reflected in the modulation of sensory and pupil responses. We hypothesized that these processes shape memory encoding, parsing the experience by grouping self- and externally generated stimuli into differentiated events. Participants encoded sound sequences, in which either the first or last few sounds were self-generated and the rest externally generated. We tested recall of the sequential order of sounds that had originated from the same (within event) or different sources (across events). Memory performance was not higher for within-event sounds, suggesting that actions did not structure the memory representation. However, during encoding, we observed the expected electrophysiological response attenuation for self-generated sounds, together with increased pupil dilation triggered by actions. Moreover, at the boundary between events, physiological responses to the first sound from the new source were influenced by the direction of the source switch. Our results suggest that introducing actions creates a stronger contextual shift than removing them, even though actions do not directly contribute to memory performance. This study contributes to our understanding of how interacting with sensory input shapes experiences by exploring the relationships between action effects on sensory responses, pupil dilation, and memory encoding. Importantly, it challenges the notion of a meaningful contribution from low-level neurophysiological mechanisms associated with action execution in the modulation of the self-generation effect.

Data from “A Registered Report Testing the Effect of Sleep on DRM False Memory: Greater Lure and Veridical Recall but Fewer Intrusions After Sleep”

This paper describes a rich dataset from a registered report investigating sleep’s effect on false memory in the Deese-Roediger-McDermott (DRM) paradigm. 534 young adults completed free recall either shortly or 12 hours after studying lists of semantic associates (e.g., hospital, nurse). Collected online, our recall data showcase high data quality, replicating classic behavioural effects (e.g., serial position curve). The dataset contains raw recall data with original spelling and recall order, accompanied by demographic information (e.g., gender, time-of-day preference). Its versatility supports reuse in modelling memory decay and search processes, understanding lexical effects and individual differences, and benchmarking online memory studies.

Large language models approach expert-level clinical knowledge and reasoning in ophthalmology: A head-to-head cross-sectional study.

Large language models (LLMs) underlie remarkable recent advanced in natural language processing, and they are beginning to be applied in clinical contexts. We aimed to evaluate the clinical potential of state-of-the-art LLMs in ophthalmology using a more robust benchmark than raw examination scores. We trialled GPT-3.5 and GPT-4 on 347 ophthalmology questions before GPT-3.5, GPT-4, PaLM 2, LLaMA, expert ophthalmologists, and doctors in training were trialled on a mock examination of 87 questions. Performance was analysed with respect to question subject and type (first order recall and higher order reasoning). Masked ophthalmologists graded the accuracy, relevance, and overall preference of GPT-3.5 and GPT-4 responses to the same questions. The performance of GPT-4 (69%) was superior to GPT-3.5 (48%), LLaMA (32%), and PaLM 2 (56%). GPT-4 compared favourably with expert ophthalmologists (median 76%, range 64-90%), ophthalmology trainees (median 59%, range 57-63%), and unspecialised junior doctors (median 43%, range 41-44%). Low agreement between LLMs and doctors reflected idiosyncratic differences in knowledge and reasoning with overall consistency across subjects and types (p>0.05). All ophthalmologists preferred GPT-4 responses over GPT-3.5 and rated the accuracy and relevance of GPT-4 as higher (p<0.05). LLMs are approaching expert-level knowledge and reasoning skills in ophthalmology. In view of the comparable or superior performance to trainee-grade ophthalmologists and unspecialised junior doctors, state-of-the-art LLMs such as GPT-4 may provide useful medical advice and assistance where access to expert ophthalmologists is limited. Clinical benchmarks provide useful assays of LLM capabilities in healthcare before clinical trials can be designed and conducted.

Emotional and temporal order effects – a comparison between word-cued and important autobiographical memories recall orders

ABSTRACT The retrieval mechanisms associated with the recall of autobiographical memories (AMs) may differ according to the cueing method used to elicit AMs. In the present study, we provide a systematic comparison between word-cued and important AMs in terms of two recall order effects, namely a temporal and an emotional order effect. The sample comprised 104 adults aged between 59 and 90 years. On two measurement occasions, participants recalled up to 20 word-cued and 15 important AMs. For each memory, participants provided their age at the time when the event occurred (age-at-event) and rated its emotionality. Order effects were analysed based on multivariate multilevel model with autoregressive effects for emotionality and age-at-event, respectively. Whereas word-cued AMs showed a stronger emotional order effect, important AMs were recalled temporarily ordered. Individuals differed in the extent to which they ordered AMs along a temporal or emotional dimension. These differences could partly be explained by personality traits. For instance, higher scores on neuroticism were associated with a stronger emotional order effect in both memory types. Findings are discussed in terms of the retrieval mechanisms that are involved in the recall of word-cued and important memory recall and how they may be affected by personality.

A Prescriptive Model for Failure Analysis in Ship Machinery Monitoring Using Generative Adversarial Networks

In recent years, advanced methods and smart solutions have been investigated for the safe, secure, and environmentally friendly operation of ships. Since data acquisition capabilities have improved, data processing has become of great importance for ship operators. In this study, we introduce a novel approach to ship machinery monitoring, employing generative adversarial networks (GANs) augmented with failure mode and effect analysis (FMEA), to address a spectrum of failure modes in diesel generators. GANs are emerging unsupervised deep learning models known for their ability to generate realistic samples that are used to amplify a number of failures within training datasets. Our model specifically targets critical failure modes, such as mechanical wear and tear on turbochargers and fuel injection system failures, which can have environmental effects, providing a comprehensive framework for anomaly detection. By integrating FMEA into our GAN model, we do not stop at detecting these failures; we also enable timely interventions and improvements in operational efficiency in the maritime industry. This methodology not only boosts the reliability of diesel generators, but also sets a precedent for prescriptive maintenance approaches in the maritime industry. The model was demonstrated with real-time data, including 33 features, gathered from a diesel generator installed on a 310,000 DWT oil tanker. The developed algorithm provides high-accuracy results, achieving 83.13% accuracy. The final model demonstrates a precision score of 36.91%, a recall score of 83.47%, and an F1 score of 51.18%. The model strikes a balance between precision and recall in order to eliminate operational drift and enables potential early action in identified positive cases. This study contributes to managing operational excellence in tanker ship fleets. Furthermore, this study could be expanded to enhance the current functionalities of engine health management software products.

A Comparative Study of the Performance of Machine Learning Models on a Tax Dataset of Yemen to Detect Levels of Tax Evasion

The performance of a classification model in machine learning is affected by many factors, such as the type of machine learning technology used. Accuracy varies from method to method. This paper presents a comparison between the performance of different models in terms of the machine learning technique used (e.g. KNN, NB, SVM, DT, RF, MLP). Based on the data provided by the Tax Authority of Yemen, which is related to the commercial and industrial profits tax, which consists of 760 attributes, after the preprocessing of data. The dataset partition technique used k-fold validation. The paper shows that the e Naïve Bayes (NB) classifier gave the highest result in accuracy and other measures. Then KNN, SVM, and RF gave the same results in accuracy 99.87%, but in SVM, KNN the results were also the same in the rest measures, while in RF models the rest measures were 97.91%,99.95%, and 98.91% in Recall, Precision and F-score in order. MLP gave 98.42 in accuracy with 66.62%, 64.21%, and 64.40 in the recall, Precision, and F-score, then DT gave 97.76% in accuracy with 57.006% ,99.24% and 72.41% in the recall, precision, and F-score.

Does telling a story in reverse elicit cues to deceit? A replication and extension of Vrij, Leal, Mann and Fisher (2012)

AbstractPurposeThe reverse order recall technique has been suggested as tool to improve deception detection accuracy. We conducted a registered replication and extension of Vrij et al., 2012's two experiments, testing whether the reverse order technique increases cues to deception in liars and accuracy in lie detection.MethodsFollowing Vrij et al., 2012, in Experiment 1, participants were interviewed twice—once lying and once telling the truth—about a mission they completed. In both interviews, participants recounted their experience in chronological and reverse order. We coded interviews for cues to deception according to a cognitive approach: those included in Vrij et al., 2012 (replication) and others included in research on the cognitive load approach (extension). In Experiment 2, participants read two transcripts (replication) or viewed two videos (extension) from Experiment 1 and decided whether senders were lying or telling the truth in continuous (replication) and dichotomous judgements (extension).ResultsIn Experiment 1, truth tellers were more detailed and plausible than liars. However, we failed to find the interaction between veracity and route recall reported by the original research on the replication or extensions cues. In Experiment 2, we only found an interaction between veracity and route recall for senders telling the truth on dichotomous lie detection judgement. However, this was not supported when examining overall accuracy.ConclusionsThese findings do not provide support for the reverse order technique as a tool to improve deception detection. We suggest further theoretical development before this technique is trained to practitioners.

Item‐based scoring and serial position of words generated during the Category Fluency Test: differences between controls and MCI subtypes

AbstractBackgroundThe order in which words are recalled during the 1‐minute Category Fluency Test (CFT) can be informative about the integrity of the underlying semantic memory (SM) system. Typically, the first entries tend to be easier exemplars and, as the category is explored in more depth during the course of the minute, words generated tend to become more difficult. We hypothesised that an item‐based scoring method that quantifies this increase in difficulty will help identify the semantic feature(s) most sensitive to the loss of SM integrity observed in people with a form of mild cognitive impairment (MCI).MethodTwo‐hundred participants (age: 58.68 ± 12.08 years) were sub‐divided into four clinical diagnostic groups: 50 healthy controls, 50 individuals with amnestic MCI, 50 individuals with non‐amnestic MCI and 50 individuals with subjective cognitive complaints but no objective deficits. No differences in age, years of education or gender existed across the four groups. All participants were administered a 1‐minute CFT (two categories: animals and fruits). Individual performance was scored word‐by‐word according to nine semantic features: age of acquisition, frequency, prevalence, recognition time, valence, arousal, dominance, sensorimotor interaction and concreteness. Correlation coefficients (Spearman’s rho) were calculated between each feature and the serial position of each word within the list. Correlations were then converted to z‐scores and analysed across the four groups with Kruskal‐Wallis one‐way analyses of variance.ResultThe z‐converted correlation between serial position and dominance was the only one differing across groups (at a p = 0.009). Bonferroni‐corrected post‐hoc tests showed a difference between healthy controls and individuals with amnestic MCI (p = 0.008). The decline in word dominance was significantly steeper in the group of amnestic MCI (z = ‐0.322) than in controls (z = ‐0.199).ConclusionThe study of SM is relevant to characterising the clinical mechanisms of normal and abnormal ageing. A procedure that focuses on CFT words difficulty and order of recall can complement classic methods that are typically used to detect loss of SM integrity. This pilot study indicates that ‘dominance’ is a semantic feature that might help characterise loss of SM integrity in individuals with amnestic MCI.

Effects of Interviewer Age on the Reminiscence Bump in Older adults’ Autobiographical Memories

ABSTRACT Introduction This study aims to investigate whether the reminiscence bump in older adults’ autobiographical memory is influenced by the interviewer’s age. Methods: I arranged four interviewer. Conditions child, university student, middle-aged adult, and older adult. The participating older adults were asked individually to describe three memories and the age at which the event occurred. I analyzed the temporal distribution of the bumps and the recall order. Results The results indicate that the interviewer’s age, especially the child interviewers were influenced the temporal location of the reminiscence bump. Conclusion: This study discusses the implications of these findings for older adults’ social communication processes.

Measure of Auditory Working Memory Span Using Monosyllabic Word Recognition Test in Young Adults With Normal Hearing: A Preliminary Study.

Measuring working memory at hearing clinics is important. This study attempted to develop a test protocol that measures auditory working memory using a standardized monosyllabic word list in Korean Speech Audiometry (KSA). We included 25 young adults with normal hearing in this study. Participants conducted word recognition and word span tests concurrently using the KSA monosyllabic word lists. We designed four test conditions according to the presence or absence of background noise and word recall order: quiet-forward (QF), quiet-backward (QB), noise-forward, (NF), and noise-backward (NB). We implemented digit span tests in the Korean Wechsler Adult Intelligence Scale-IV (K-WAIS-IV) to determine the validity of the working memory outcomes. Word recognition scores of QF and QB were significantly higher than those of NF and NB. The percentages of correctly recalled words and word recall span scores were highest in QF and lowest in NB. Overall, the Pearson correlation and multiple regression statistics showed that our word span test outcomes for QB and NF were highly associated with digit span scores on the K-WAIS-IV. Our proposed test protocol showed the possibility of measuring auditory working memory and monosyllabic word recognition simultaneously by validating the results with K-WAIS-IV outcomes.

Forward and backward recall dynamics.

Although possible to recall in both forward and backward order, recall proceeds most naturally in the order of encoding. Prior studies ask whether and how forward and backward recall differ. We reexamine this classic question by studying recall dynamics while varying the predictability and timing of forward and backward cues. Although overall accuracy did not differ by recall direction, recall dynamics highlight key distinctions. Forward recall exhibits a modest advantage for correct transitions following errors, independent of cueing predictability and list length. Without consistent directional cueing, participants initiate backward recall more accurately, but this effect reverses with predictable directional cues. Following omissions, participants commit more fill-in errors in backward recall. Our findings implicate an asymmetric, cue-dependent retrieval process underlying forward and backward recall, with relative contributions of primacy and recency depending on directional predictability. (PsycInfo Database Record (c) 2023 APA, all rights reserved).

Sequential syntactic knowledge supports item but not order recall in verbal working memory.

Previous studies have shown that psycholinguistic effects such as lexico-semantic knowledge effects mainly determine item recall in verbal working memory (WM). However, we may expect that syntactic knowledge, involving knowledge about word-level sequential aspects of language, should also impact serial-order aspects of recall in WM. Evidence for this assumption is scarce and inconsistent and has been conducted in language with deterministic syntactic rules. In languages such as French, word position is determined in a probabilistic manner: an adjective is placed before or after a noun, depending on its lexico-semantic properties. We exploited this specificity of the French language for examining the impact of syntactic positional knowledge on both item and serial order recall in verbal WM. We presented lists with adjective-noun pairs for immediate serial recall, the adjectives being in regular or irregular position relative to the nouns. We observed increased recall performance when adjectives occurred in regular position; this effect was observed for item recall but not order recall scores. We propose an integration of verbal WM and syntactic processing models to account for this finding by assuming that the impact of syntactic knowledge on serial-order WM recall is indirect and mediated via syntax-dependent item-retrieval processes.

Diagnosis of patellofemoral osteoarthritis using enhanced sequential deep learning techniques

The surface electromyography (sEMG) signal is a complex interference pattern resulting from the electrical activity of contracting muscles, which directly correlates with muscle activity and exercise level. Patellofemoral osteoarthritis (PF OA) refers to the softening and destruction of the articular cartilage of the knee cap, which is important to diagnose in the early phase. The aim of this study is to develop a predictive model, called the Enhanced Sequencing Deep Learning (PESDL) Algorithm, to detect PF OA from the sEMG signal. The algorithm consists of four main modules: (i) Data Acquisition Module (DAM), (ii) Signal Preprocessing Module (SPM), (iii) Data Augmentation and Concatenation Module (DACM), and (iv) Lower Limb Classification (LLC). The sEMG signals from five core muscles of the lower limb were acquired during stepping stairs up and down, and machine learning was used to obtain the muscle activation data signals. The acquired data was preprocessed, augmented, concatenated, and shuffled. Three feedforward deep networks (RNN, LSTM, and GRU) were used to classify the lower limb sEMG time-sequence data, with the GRU network showing better performance than the other two. The model's performance was evaluated using seven performance measures and 10-fold cross-validation. The maximum values of feedforward deep networks (RNN), LSTM, and GRU for the five muscles (RF, BF, VM, ST, and FX) using seven performance measures were as follows: RNN (0.961, 0.154, 0.943, 0.244, 1.0, 1.0, 0.899), LSTM (0.967, 0.147, 0.945, 0.217, 0.997, 1.0, 0.927), and GRU (0.976, 0.111, 0.949, 0.212, 0.996, 0.998, 0.938), respectively, listed in the order of accuracy, loss, validation accuracy, validation loss, recall, precision, and F1-score. Comparisons with other state-of-the-art models using the same datasets demonstrated the effective performance of the predictive model. The results suggest that RNN models, particularly GRU and LSTM, can be effective for detecting PF OA, and the specific RNN architecture used can have a significant impact on performance.

How Music Moves Us: Music-induced Emotion Influences Motor Learning

Music is an important tool for the induction and regulation of emotion. Although learning a sequential motor behaviour is essential to normal motor function, to our knowledge, the role of music-induced emotion on motor learning has not been explored. Our experiment aimed to determine whether listening to different emotional music could influence motor sequence learning. We focused on two sub-components of motor sequence learning: the acquisition of the order of the elements in the sequence (the “what”), and the ability to carry out the sequence, combining the elements in a single, skilled action (the “how”). Twenty subjects performed a motor sequence-learning task with a digitizing tablet in three different experimental sessions. In each session they executed the task while listening to three different musical pieces, eliciting fearful, pleasant, and neutral mood. Eight targets were presented in a pre-set order and subjects were asked to learn the sequence while moving. Music-induced pleasure had an impact on movement kinematics with onset time and peak velocity decreasing and movement time increasing more with respect to neutral music session. Declarative learning, verbal recall of the sequence order, was improved under emotional manipulation, but only for fear-condition. Results suggest that music-induced emotion can influence both sub-components of motor learning in a different way. Music-induced pleasure may have improved motor components of sequence learning by means of increased striatal dopamine availability whereas music-induced fear may facilitate the recruitment of attentional circuits, thus acting on declarative knowledge of the sequence order.

Neural temporal context reinstatement of event structure during memory recall.

The transformation of experiences into meaningful events and memories is intertwined with the notion of time. Temporal perception can influence, and be influenced by, segmenting continuous experience into meaningful events. Episodic memories formed from these events become associated with temporal information as well. However, it is less clear how temporal perception contributes to structuring events and organizing memory: whether it plays a more active or passive role, and whether this temporal information is encoded initially during perception or influenced by retrieval processes. To address these questions, we examined how event segmentation influences temporal representations during initial perception and memory retrieval, without testing temporal information explicitly. Using a neural measure of temporal context extracted from scalp electroencephalography in human participants (N = 170), we found reduced temporal context similarity between studied items separated by an event boundary when compared to items from the same event. Furthermore, while participants freely recalled list items, neural activity reflected reinstatement of temporal context representations from the study phase, including temporal disruption. A computational model of episodic memory, the context maintenance and retrieval (CMR) model, predicted these results, and made novel predictions regarding the influence of temporal disruption on recall order. These findings implicate the impact of event structure on memory organization via temporal representations, underscoring the role of temporal information in event segmentation and episodic memory. (PsycInfo Database Record (c) 2023 APA, all rights reserved).

Differences between healthy adults and mild cognitive impairment patients in Category Fluency performance: A scoring method based on words’ Serial Recall Order and item‐level features

AbstractBackgroundSemantic Memory (SM) alterations occur very early in preclinical AD. Category Fluency (CF) is the most widely‐used measure of free SM recall but is not effective at detecting subtle SM changes. Its score based on word‐count, in fact, is also influenced by other abilities such as processing speed and executive functioning.We thus designed an alternative, item‐level scoring method less susceptible to these intervenient variables. This quantifies “how difficult” each CF word is based on 9 semantic features. Additionally, the Serial Recall Order (SRO) is established. This is an ordinal variable that tags the position of each word within the recalled list. Participant‐specific correlations are then calculated between SRO and each semantic feature. These capture the tendency of each participant to recall increasingly difficult words.We previously showed that, in healthy adults, these correlational indices are associated with neurocognitive mechanisms of semantic processing and control, but are statistically independent of processing speed and executive functioning (https://doi.org/10.3389/fnagi.2021.678588; https://doi.org/10.3389/fpsyg.2021.777838).This study tests whether these indices differ between normal adults and patients with mild cognitive impairment (MCI) at risk of progressing to AD.MethodCF performance from 60 MCI patients and 60 matched controls was scored following the above methodology. ANCOVAs were run to test the effect of diagnosis on z‐converted correlational indices, controlling for age, educational attainment, and number of CF entries at the basis of the correlations.ResultOf the nine indices, ‘SRO‐dominance’ was the only one differing between diagnoses. This correlation was ≈‐0.204 in controls and ≈‐0.362 in patients. Words 1‐8 were of similar dominance between groups. A difference was found instead for Words 9‐12.Conclusion‘Dominance’ is the ‘degree of control/power’ conveyed by a word. While the general tendency of CF performance is that of recalling words that are increasingly dominant, this trend is significantly more pronounced in MCI. This suggests a different use of semantic‐control resources between groups when the target category is explored in more depth. Although preliminary, these findings indicate that the SRO of CF words can help profile SM in patients with suspected AD. More work is needed, however, to study diagnostic statuses informed by biomarkers.

Differential influence of first- vs. third-person visual perspectives on segmentation and memory of complex dynamic events

Until now, most studies investigating the relationship between event segmentation and memory have used videos filmed from a third-person perspective, although people experience their lives from a first-person perspective. The present study aimed to determine whether visual perspective impacts events segmentation and further recall. Fifty-seven participants were recruited and assigned to either first- (1PP) or third-person perspective (3PP) condition, before segmenting videos of daily life activities. Our results showed that the although the number of event boundaries was higher in the 3PP condition than in the 1PP, no differences were observed for event segmentation qualitative abilities and organization. Memory of temporal order was better for events encoded in the 3PP than in the 1PP, while memory content was similar in both conditions. Higher event segmentation rates were correlated with a better recall of small actions and temporal order.

An acoustic detection dataset of birds (Aves) in montane forests using a deep learning approach.

Long-term monitoring is needed to understand the statuses and trends of wildlife communities in montane forests, such as those in Yushan National Park (YSNP), Taiwan. Integrating passive acoustic monitoring (PAM) with an automated sound identifier, a long-term biodiversity monitoring project containing six PAM stations, was launched in YSNP in January 2020 and is currently ongoing. SILIC, an automated wildlife sound identification model, was used to extract sounds and species information from the recordings collected. Animal vocal activity can reflect their breeding status, behaviour, population, movement and distribution, which may be affected by factors, such as habitat loss, climate change and human activity. This massive amount of wildlife vocalisation dataset can provide essential information for the National Park's headquarters on resource management and decision-making. It can also be valuable for those studying the effects of climate change on animal distribution and behaviour at a regional or global scale. To our best knowledge, this is the first open-access dataset with species occurrence data extracted from sounds in soundscape recordings by artificial intelligence. We obtained seven bird species for the first release, with more bird species and other taxa, such as mammals and frogs, to be updated annually. Raw recordings containing over 1.7 million one-minute recordings collected between the years 2020 and 2021 were analysed and SILIC identified 6,243,820 vocalisations of seven bird species in 439,275 recordings. The automatic detection had a precision of 0.95 and the recall ranged from 0.48 to 0.80. In terms of the balance between precision and recall, we prioritised increasing precision over recall in order to minimise false positive detections. In this dataset, we summarised the count of vocalisations detected per sound class per recording which resulted in 802,670 occurrence records. Unlike data from traditional human observation methods, the number of observations in the Darwin Core "organismQuantity" column refers to the number of vocalisations detected for a specific bird species and cannot be directly linked to the number of individuals.We expect our dataset will be able to help fill the data gaps of fine-scale avian temporal activity patterns in montane forests and contribute to studies concerning the impacts of climate change on montane forest ecosystems on regional or global scales.