Task Performance Research Articles

Does innovative behaviour intervene between budgetary participation and performance in the public sector?

Purpose For a public servant (PS) to be innovative, he or she needs to gather and process enough vital information from budget setting processes. However, research addressing how budgetary participation (BP) can trigger innovative behaviour (IB) in PSs and eventually foster task performance (TP) is rare, which is why the authors conduct this research. The purpose of this study is to understand how BP shapes TP through the IB of PSs. Design/methodology/approach The authors develop and test a mediation model with 860 responses from public sector workers across 25 government agencies using the PLS-SEM technique of Smart PLS 4. Possible control factors were addressed for both the mediator and target-independent construct. In particular, the authors use sex, age and tenure as control factors for IB. Also, the authors use job satisfaction, job engagement and perceived fairness in the budgetary system as control factors for TP. Findings The authors find a favourable and significant relationship between BP and TP; BP and IB; and IB and TP. The authors also find that IB partially mediates the relationship between BP and TP, such that BP fosters TP through the innovativeness of PSs. This finding suggests that PSs who participate in budget preparation are able to innovate, which, in turn enable them to perform tasks effectively. Research limitations/implications The authors call on forthcoming researchers to test the mediation model in other public sector settings worldwide. They may also consider other variables that can possibly mediate the positive impacts of BP on TP. Practical implications Lessons are discussed for governments, human resources directors and managers, management accountants, budget officers, procurement officers and other public sector workers and consultants. Originality/value The authors show how BP fosters TP through the innovativeness of PSs, since there is much more to know in this regard. The authors also help to resolve the paradox of inconsistency in the BP–TP literature by using IB as a mediator.

Dynamic Graph Guided Progressive Partial View-Aligned Clustering.

In recent years, there has been a growing focus on multiview data, driven by its rich complementary and consistent information, which has the potential to significantly enhance the performance of downstream tasks. Although many multiview clustering (MVC) methods have achieved promising results by integrating the information of multiple views to learn the consistent representation or consistent graph, these methods typically require complete and entirely accurate correspondences between multiview data, which is challenging to fulfill in practice leading to the problem of partially view-aligned clustering (PVC). To tackle it, we propose a novel method, called dynamic graph guided progressive partial view-aligned clustering (DGPPVC) in this article. To the best of our knowledge, this could be the first work to employ graph convolutional network (GCN) to address the problem of PVC, which explores GCN with dynamic adjacency matrix to reduce unreliable alignments and locate the feature representation with consistent graph structure. In particular, DGPPVC develops an end-to-end framework that encompasses graph construction, feature representation learning, and alignment relationships learning, in which the three parts mutually influence and benefit each other. Moreover, DGPPVC adopts a novel alignment learning strategy that progresses from simplicity to complexity, enabling the step-by-step acquisition of unknown correspondences between different modalities. By giving priority to simple instance pairs, a variant of Jaccard similarities is designed to identify more reliable and complex alignments progressively. During the gradual learning process of alignment relationships, the graph structure matrix is continually and dynamically optimized, thus acquiring a greater variety of graph information between different views. Experiments on several real-world datasets show our promising performance compared with the state-of-the-art methods in partially view-aligned clustering.

Task reliability index for operator performance and failure probability assessment in control room simulators

Main control room simulators are an important information source for improving the empirical basis of human reliability analysis (HRA). Both quantitative and qualitative observations provide valuable insight on the operators’ ability to successfully perform their tasks. However, relevant portions of this evidence are not used in typical approaches for estimation of human error probabilities (HEP), which are based on failure counts. With small sample sizes and the operator performance levels representative for nuclear power plants, very little information enters the probability estimates. This paper presents a novel approach to formalize the collection and use of reliability-relevant evidence from simulators. The methodology treats task reliability with the plant-centred orientation characteristic of the systems or Probabilistic Safety Assessment (PSA) perspective as well as a human-centred view of task performance based in human factors engineering. These are combined to form a Task reliability index (TRI) able to distinguish finer performance variations and consider indications of potential failures. Bayesian updating of HEP distributions is suggested by translating the TRI data into pseudo-failures and trials, preserving the convenient coupling of Beta-Binomial distributions in the Bayesian analysis. The approach is demonstrated using data from the International HRA Empirical Study and is shown to produce results that are comparable to those obtained in the empirical study, which were based on comprehensive expert analyses.

RevGNN: Negative Sampling Enhanced Contrastive Graph Learning for Academic Reviewer Recommendation

Acquiring reviewers for academic submissions is a challenging recommendation scenario. Recent graph learning-driven models have made remarkable progress in the field of recommendation, but their performance in the academic reviewer recommendation task may suffer from a significant false negative issue. This arises from the assumption that unobserved edges represent negative samples. In fact, the mechanism of anonymous review results in inadequate exposure of interactions between reviewers and submissions, leading to a higher number of unobserved interactions compared to those caused by reviewers declining to participate. Therefore, investigating how to better comprehend the negative labeling of unobserved interactions in academic reviewer recommendations is a significant challenge. This study aims to tackle the ambiguous nature of unobserved interactions in academic reviewer recommendations. Specifically, we propose an unsupervised Pseudo Neg-Label strategy to enhance graph contrastive learning (GCL) for recommending reviewers for academic submissions, which we call RevGNN. RevGNN utilizes a two-stage encoder structure that encodes both scientific knowledge and behavior using Pseudo Neg-Label to approximate review preference. Extensive experiments on three real-world datasets demonstrate that RevGNN outperforms all baselines across four metrics. Additionally, detailed further analyses confirm the effectiveness of each component in RevGNN.

TCohPrompt: task-coherent prompt-oriented fine-tuning for relation extraction

AbstractPrompt-tuning has emerged as a promising approach for improving the performance of classification tasks by converting them into masked language modeling problems through the insertion of text templates. Despite its considerable success, applying this approach to relation extraction is challenging. Predicting the relation, often expressed as a specific word or phrase between two entities, usually requires creating mappings from these terms to an existing lexicon and introducing extra learnable parameters. This can lead to a decrease in coherence between the pre-training task and fine-tuning. To address this issue, we propose a novel method for prompt-tuning in relation extraction, aiming to enhance the coherence between fine-tuning and pre-training tasks. Specifically, we avoid the need for a suitable relation word by converting the relation into relational semantic keywords, which are representative phrases that encapsulate the essence of the relation. Moreover, we employ a composite loss function that optimizes the model at both token and relation levels. Our approach incorporates the masked language modeling (MLM) loss and the entity pair constraint loss for predicted tokens. For relation level optimization, we use both the cross-entropy loss and TransE. Extensive experimental results on four datasets demonstrate that our method significantly improves performance in relation extraction tasks. The results show an average improvement of approximately 1.6 points in F1 metrics compared to the current state-of-the-art model. Codes are released at https://github.com/12138yx/TCohPrompt.

Performance Tasks in EFL Classes: Massive Open Online Courses (MOOCs)

This study aims to investigate the integration of massive open online courses (MOOCs) as performance tasks at the K12 level in EFL classes during COVID-19. A pre/post survey including open-ended questions was developed for pilot testing, 110 K-12 students in Turkey in the 2020-2021 school year during the spread of Covid-19 completed these surveys. Both qualitative and quantitative data were used in data collection. Statistical analysis was used in the analysis of the quantitative data, and descriptive and content analysis was used in the analysis of qualitative data. It is thought that the results of the research will guide researchers and practitioners in the theoretical and practical aspects of online and blended learning at the K12 level.

University students’ information literacy skills and impact of training in Bangladesh

This article examines university students’ perceived skill levels in utilizing internet resources and library tools, as well as accomplishing specific information-skills tasks, and assesses the impact of information literacy training in Bangladesh. The study involved two phases: first, a survey was administered to a sample of 257 students and, second, an experiment was conducted with 28 students to examine the effectiveness of training. The survey results show that the students perceived their skills with internet resources and library tools to be at a high level, with significant differences found in skill levels based on gender and age. Additionally, the students reported a high level of perceived skills in completing information tasks, with no significant differences found across the demographic variables. In the experimental study, the participants exhibited low task performance on a set of 22 task questions prior to training, achieving an overall correct score of 28.4%. However, a significant improvement in correct scores (57.1%) was observed after participating in the skills training.

The Performance of Kaizen Tasks Across Three Online Discrete Choice Experiment Surveys: An Evidence Synthesis.

Kaizen is a Japanese term for continuous improvement (kai ~ change, zen ~ good). In a kaizen task, a respondent makes sequential choices to improve an object's profile, revealing a preference path. Including kaizen tasks in a discrete choice experiment has the advantage of collecting greater preference evidence than pick-one tasks, such as paired comparisons. OBJECTIVE ANDMETHODS: So far, three online discrete choice experiments have included kaizen tasks: the 2020 US COVID-19 vaccination (CVP) study, the 2021 UK Children's Surgery Outcome Reporting (CSOR) study, and the 2023 US EQ-5D-Y-3L valuation (Y-3L) study. In this evidence synthesis, we describe the performance of the kaizen tasks in terms of response behaviors, conditional logit and Zermelo-Bradley-Terry (ZBT) estimates, and their standard errors in each of the surveys. Comparing the CVP and Y-3L, including hold-outs (i.e., attributes shared by all alternatives) seems to reduce positional behavior by half. The CVP tasks excluded multi-level improvements; therefore, we could not estimate logit main effects directly. In the CSOR, only 12 of the 21 logit estimates are significantly positive (p < 0.05), possibly due to the fixed attribute order. All Y-3L estimates are significantly positive, and their predictions are highly correlated (Pearson: logit 0.802, ZBT 0.882) and strongly agree (Lin: logit 0.744, ZBT 0.852) with the paired-comparison probabilities. These discrete choice experiments offer important lessons for future studies: (1) include warm-up tasks, hold-outs, and multi-level improvements; (2) randomize the attribute order (i.e., up-down) at the respondent level; and (3) recruit smaller samples of respondents than traditional discrete choice experiments with only pick-one tasks.

Leader affiliative humour and work outcomes in multicultural teams

ABSTRACT Leader affiliative humour is when the leader makes funny remarks to reduce tension or amuse followers. Previous authors find leader affiliative humour may produce favourable effects. Yet, research demonstrates that it can also cause undesirable results. This study draws on the cultural perspective to propose follower power distance orientation as a moderator in leader affiliative humour’s favourable and unfavourable influences on work outcomes in multicultural teams, which refers to work teams composed of members from different countries and cultural backgrounds. To collect data from members of multicultural work teams, we conducted an individual-level three-wave survey with a two-month time lag each in a large Taiwan-based financial holding company operating in more than 20 countries. Analyses of matched data collected from 192 worker-supervisor pairs find leader affiliative humour can improve task performance via increased leader-member exchange only when the follower’s power distance is low and can initiate organizational deviant behaviours via perceived acceptability of norm violation only when the follower’s power distance is high. Our study provides empirical evidence from multicultural teams to demonstrate how leader affiliative humour may favourable or unfavourable influence workers with different levels of power distance orientation to complement the leader humour and the multicultural team literature.

Pre-movement muscle co-contraction associated with motor performance deterioration under high reward conditions

Reward usually enhances task performance, but exceptionally large rewards can impede performance due to psychological pressure. In this study, we investigated motor activity changes in high-reward situations and identified indicators for performance decline. Fourteen healthy adults practiced a velocity-dependent right-hand motor task for three days, followed by a test day with varying monetary reward for each trial. Participants were divided into low performers (LPs) and high performers (HPs) according to whether success rate decreased or increased, respectively, on the highest reward trials compared to lower reward trials. Both LPs and HPs demonstrated increased hand velocity during higher reward trials, but only LPs exhibited a significant increase in velocity variance. There was also a negative correlation between the pre-movement co-contraction index (CCI) of the biceps and triceps muscles and success rate on the highest reward trials. This correlation was confirmed in a second experiment with 12 newly recruited participants, suggesting that pre-movement CCI is a marker for performance decline caused by high reward. These findings suggest that interventions to reduce pre-movement CCI such as biofeedback training could be useful for preventing the paradoxical decline in motor performance associated with high rewards.

Thriving in change: examining the influence of organizational sustainability on employee performance in the post-COVID landscape

Purpose The economic, financial and social crisis has heightened researchers' interest in the importance of organizational sustainability and resources (e.g. work engagement and psychological capital) for performance. Despite some studies demonstrating the significance of resources, such as work engagement and psychological capital, for workers’ performance, there is limited knowledge regarding the role of organizational sustainability practices (OSP) in the relationship mentioned above. Therefore, this study drew upon the social identity theory to conceptualize and test the mediating role of work engagement in the relationship between OSP and performance (both task and adaptive). Based on the conservation of resources theory, it was hypothesized that a worker's resource – psychological capital – would moderate the indirect relationship between social responsibility practices and performance (both task and adaptive). Design/methodology/approach To test the proposed model, 304 adults participated in a two-wave study. Findings The results supported the hypotheses, revealing that OSP had a positive and significant relationship with task and adaptive performance, occurring through work engagement. Furthermore, the indirect relationship between OSP and performance task and adaptive through work engagement was moderated by psychological capital, such that the relationship became stronger for workers with lower levels of psychological capital. Originality/value The relevance of this research lies in its ability to identify the conditions and mechanisms in which corporate social responsibility initiatives are most effective, thereby enabling a more nuanced application of these practices across different sectors and organizational cultures. This study's theoretical and practical implications may assist organizations in devising strategies for implementing social responsibility practices to enhance their outcomes.

Robust Brain Correlates of Cognitive Performance in Psychosis and its Prodrome

BackgroundNeurocognitive impairment is a well-known phenomenon in schizophrenia that begins prior to psychosis onset. Connectome-wide association studies have inconsistently linked cognitive performance to resting-state fMRI. We hypothesized a carefully selected cognitive instrument and refined population would allow identification of reliable brain-behavior associations with connectome-wide association studies. To test this hypothesis, we first identified brain-cognition correlations via a connectome-wide association study in early psychosis. We then asked, in an independent dataset, if these brain-cognition relationships would generalize to individuals who develop psychosis in the future. MethodsThe Seidman Auditory Continuous Performance Task (ACPT) effectively differentiates healthy participants from those with psychosis. Our connectome-wide association study used the Human Connectome Project for Early Psychosis (n=183) to identify links between connectivity and ACPT performance. We then analyzed the North American Prodrome Longitudinal Study 2 (n=345), a multi-site prospective study of individuals at risk for psychosis. We tested the connectome-wide association study-identified cognition-connectivity relationship in both individuals at risk for psychosis and controls. ResultsOur connectome-wide association study in early-course psychosis identified robust associations between better ACPT performance and higher prefrontal-somatomotor connectivity (p<.005). Prefrontal-somatomotor connectivity was also related to ACPT performance in at-risk individuals who would develop psychosis (n=17). This finding was not observed in nonconverters (n=196) or controls (n=132). ConclusionsThis connectome-wide association study identified reproducible links between connectivity and cognition in separate samples of psychosis and at-risk individuals who would later develop psychosis. A carefully selected task and population improves the ability of connectome-wide association studies to identify reliable brain-phenotype relationships.

Contributions of cortical neuron firing patterns, synaptic connectivity, and plasticity to task performance

Neuronal responses during behavior are diverse, ranging from highly reliable ‘classical’ responses to irregular ‘non-classically responsive’ firing. While a continuum of response properties is observed across neural systems, little is known about the synaptic origins and contributions of diverse responses to network function, perception, and behavior. To capture the heterogeneous responses measured from auditory cortex of rodents performing a frequency recognition task, we use a novel task-performing spiking recurrent neural network incorporating spike-timing-dependent plasticity. Reliable and irregular units contribute differentially to task performance via output and recurrent connections, respectively. Excitatory plasticity shifts the response distribution while inhibition constrains its diversity. Together both improve task performance with full network engagement. The same local patterns of synaptic inputs predict spiking response properties of network units and auditory cortical neurons from in vivo whole-cell recordings during behavior. Thus, diverse neural responses contribute to network function and emerge from synaptic plasticity rules.

Calculation Model for the Exit Decision Sight Distance of Right-Turn Ramps on the Left at Interchange

Given the rapid construction of freeways in developing countries such as China, land use is constantly under strict constraints, leading to challenges in adopting conventional layouts for interchanges. Implementing right-turn ramps on the left (RTRL) at interchanges can minimize land occupancy; however, the traffic safety level in this type of diversion area design requires extra attention. This study examines the decision sight distance for right-turn exit ramps on the left side. Utilizing unmanned aerial vehicle (UAV) video and the YOLOv3 target detection algorithm, the original trajectory data of vehicles in the diversion area is extracted. Employing Kalman filtering and Frenet coordinate system conversion reveals microscopic vehicle lane-change patterns, velocities, and time headways. Furthermore, the driving simulation experiment assesses driver behaviors in RTRL, with subjective, task performance, and physiological measure indicators. Ultimately, the range of the decision sight distance is defined, and establishing a calculation model involves determining relevant parameters based on measured data and simulation outcomes. The results indicate potential insufficiencies in the decision sight distance when standardized values are applied to RTRL.

Spatial Contextual Information Modulates Affordance Processing and Early Electrophysiological Markers of Scene Perception.

Scene perception allows humans to extract information from their environment and plan navigation efficiently. The automatic extraction of potential paths in a scene, also referred to as navigational affordances, is supported by scene-selective regions (SSRs) that enable efficient human navigation. Recent evidence suggests that the activity of these SSRs can be influenced by information from adjacent spatial memory areas. However, it remains unexplored how these contextual information could influence the extraction of bottom-up information, such as navigational affordances, from a scene and the underlying neural dynamics. Therefore, we analyzed ERPs in 26 young adults performing scene and spatial memory tasks in artificially generated rooms with varying numbers and locations of available doorways. We found that increasing the number of navigational affordances only impaired performance in the spatial memory task. ERP results showed a similar pattern of activity for both tasks, but with increased P2 amplitude in the spatial memory task compared with the scene memory. Finally, we reported no modulation of the P2 component by the number of affordances in either task. This modulation of early markers of visual processing suggests that the dynamics of SSR activity are influenced by a priori knowledge, with increased amplitude when participants have more contextual information about the perceived scene. Overall, our results suggest that prior spatial knowledge about the scene, such as the location of a goal, modulates early cortical activity associated with SSRs, and that this information may interact with bottom-up processing of scene content, such as navigational affordances.

A dynamic graph structural framework for implicit sentiment identification based on complementary semantic and structural information

Implicit sentiment identification has become the classic challenge in text mining due to its lack of sentiment words. Recently, graph neural network (GNN) has made great progress in natural language processing (NLP) because of its powerful feature capture ability, but there are still two problems with the current method. On the one hand, the graph structure constructed for implicit sentiment text is relatively single, without comprehensively considering the information of the text, and it is more difficult to understand the semantics. On the other hand, the constructed initial static graph structure is more dependent on human labor and domain expertise, and the introduced errors cannot be corrected. To solve these problems, we introduce a dynamic graph structure framework (SIF) based on the complementarity of semantic and structural information. Specifically, for the first problem, SIF integrates the semantic and structural information of the text, and constructs two graph structures, structural information graph and semantic information graph, respectively, based on specialized knowledge, which complements the information between the two graph structures, provides rich semantic features for the downstream identification task, and helps to understanding of the contextual information between implicit sentiment semantics. To deal with the second issue, SIF dynamically learns the initial static graph structure to eliminate the noise information in the graph structure, preventing noise accumulation that affects the performance of the downstream identification task. We compare SIF with mainstream natural language processing methods in three publicly available datasets, all of which outperform the benchmark model. The accuracy on the Puns of day dataset, SemEval-2021 task 7 dataset, and Reddit dataset reaches 95.73%, 85.37%, and 65.36%, respectively. The experimental results demonstrate a good application scenario for our proposed method on implicit sentiment identification tasks.

MICK (Mobile Integrated Cognitive Kit) App for Concussion Assessment in a Youth Ice Hockey League.

Visual symptoms are common after concussion. Rapid automatized naming (RAN) tasks are simple performance measures that demonstrate worse time scores in the setting of acute or more remote injury. We evaluated the capacity for the Mobile Universal Lexicon Evaluation System (MULES) and Staggered Uneven Number (SUN) testing to be feasibly administered during preseason testing in a cohort of youth ice hockey athletes using a novel computerized app, the Mobile Integrated Cognitive Kit (MICK). Participants from a youth hockey league underwent preseason testing. Among 60 participants, the median age was 13 years (range 6-17). The median best time for the MULES was 49.8 seconds (range = 34.2-141.0) and the median best time for the SUN was 70.1 (range = 36.6-200.0). As is characteristic of timed performance measures, there were learning effects between the first and second trials for both the MULES (median improvement = 10.6 seconds, range = -32.3 to 92.0, P < 0.001, Wilcoxon signed-rank test) and SUN (median improvement = 2.4 seconds, range= -8.0 to 15.1, P = 0.001, Wilcoxon signed-rank test). Age was a predictor of best baseline times, with longer (worse) times for younger participants for MULES (P < 0.001, rs = -0.67) and SUN (P < 0.001, rs = -0.54 Spearman rank correlation). Degrees of learning effect did not vary with age (P > 0.05, rs = -0.2). Vision-based RAN tasks, such as the MULES and SUN, can be feasibly administered using the MICK app during preseason baseline testing in youth sports teams. The results suggest that more frequent baseline tests are necessary for preadolescent athletes because of the relation of RAN task performance to age.

Acute inorganic nitrate ingestion does not impact oral microbial composition, cognitive function, or high-intensity exercise performance in female team-sport athletes.

The purpose of this study was to investigate the effects of acute nitrate (NO3-)-rich beetroot juice ingestion on explosive and high-intensity exercise performance, oral microbiota composition, and cognitive flexibility (i.e., function), before and after maximal intermittent running exercise. Fifteen women team-sport athletes were assigned in a randomized, double-blind, crossover design to consume concentrated NO3--depleted beetroot juice (PL; 0.1mmol NO3-) and NO3--rich beetroot juice (BR; 12.0mmol NO3-) 2.5h prior to performing a battery of exercise performance tasks and cognitive testing before and after the Yo-Yo intermittent recovery level 1 (YYIR1) running test. Resting plasma [NO3-] and plasma nitrite ([NO2-]) were elevated following BR (P < 0.001). BR did not impact global composition or relative abundance of taxa in the oral microbiome (P > 0.05) or cognitive flexibility before or after exercise (P > 0.05). There was no significant difference in performance during 20-m (PRE, PL: 4.38 ± 0.27 vs. BR: 4.38 ± 0.32s; POST, PL: 4.45 ± 0.29 vs. BR: 4.43 ± 0.35s) and 10-m sprints (PRE, PL 2.78 ± 0.15 vs. BR 2.79 ± 0.18s; POST, PL: 2.82 ± 0.16 vs. BR: 2.81 ± 0.19s), isokinetic handgrip dynamometry, medicine ball throw, horizontal countermovement jump, or YYIR1 (PL: 355 ± 163m vs. BR: 368 ± 184m) between BR and PL (P > 0.05). These findings indicate that acute dietary NO3- may not influence the oral microbiome, explosive and high-intensity exercise performance, or cognitive function in women team-sport athletes.

Learning analytics dashboard design: Workplace learner preferences for reference frames in immersive training in practice

AbstractBackgroundLearning analytics dashboards are increasingly being used to communicate feedback to learners. However, little is known about learner preferences for dashboard designs and how they differ depending on the self‐regulated learning (SRL) phases the dashboards are presented (i.e., forethought, performance, and self‐reflection phases) and SRL skills. Insight into design preferences for dashboards with different reference frames (i.e., progress, social, internal achievement and external achievement) is important because the effectiveness of feedback can depend upon how a learner perceives it.ObjectiveThis study examines workplace learner preferences for four dashboard designs for each SRL phase and how SRL skills relate to these preferences.MethodsSeventy participants enrolled in a chemical process apprenticeship program took part in the study. Preferences were determined using a method of adaptive comparative judgement and SRL skills were measured using a questionnaire. Preferences were tested on four dashboard designs informed by social and temporal comparison theory and goal setting theory. Multinomial logistic regressions were used to examine the relationship between dashboard preferences and SRL.Results and ConclusionsResults show that the progress reference frame is more preferred before and after task performance, and the social reference frame is less preferred before and after task performance. It was found that the higher the SRL skill score the higher the probability a learner preferred the progress reference frame compared to having no preference before task performance. The results are consistent with other findings, which suggest caution when using social comparison in designing dashboards which provide feedback.

An improved efficientnet B7 underwater marine organism image classification model based on the spatial group enhancement attention mechanisms

In this paper, we adopt the Efficientnet B7 model, which is an excellent performer in deep learning algorithms, and improve it by combining with the Spatial Group Enhancement attention mechanism to provide an innovative method for underwater biological image classification. After a reasonable division of the dataset, we introduced the improved Efficientnet B7 model for training. By observing the change of the loss curve on the training set, it can be seen that the loss value of the model gradually decreases from 0.6 to 0.05 and tends to be stable, which indicates that the classification prediction of the model is gradually close to the real result, and it can classify the underwater marine organisms images effectively. The accuracy on the training set gradually increases from the initial 80.12% to 98.31%, and the recall also grows from 75.82% to 98.93%, showing that the model predicts well. The performance on the test set shows an accuracy of 84.86%, a recall of 82.40%, and an F1 value of 80.2%, indicating that the model has good generalisation ability. This means that our proposed improved methods and ideas have achieved significant results in the task of underwater biological image classification, providing a useful reference for research in related fields. By comparing the experimental results, it can be seen that the optimisation of the Efficientnet B7 model by introducing the Spatial Group Enhancement attention mechanism achieves better performance in the underwater marine organisms image classification task, which lays the foundation for the future application of deep learning in underwater environments. This study is of positive significance for the promotion of underwater biology research, marine protection and other fields, and provides new ideas and methods for carrying out more in-depth related research.