Diverse Items Research Articles

Diversifying Collaborative Filtering via Graph Spreading Network and Selective Sampling.

Graph neural network (GNN) is a robust model for processing non-Euclidean data, such as graphs, by extracting structural information and learning high-level representations. GNN has achieved state-of-the-art recommendation performance on collaborative filtering (CF) for accuracy. Nevertheless, the diversity of the recommendations has not received good attention. Existing work using GNN for recommendation suffers from the accuracy-diversity dilemma, where slightly increases diversity while accuracy drops significantly. Furthermore, GNN-based recommendation models lack the flexibility to adapt to different scenarios' demands concerning the accuracy-diversity ratio of their recommendation lists. In this work, we endeavor to address the above problems from the perspective of aggregate diversity, which modifies the propagation rule and develops a new sampling strategy. We propose graph spreading network (GSN), a novel model that leverages only neighborhood aggregation for CF. Specifically, GSN learns user and item embeddings by propagating them over the graph structure, utilizing both diversity-oriented and accuracy-oriented aggregations. The final representations are obtained by taking the weighted sum of the embeddings learned at all layers. We also present a new sampling strategy that selects potentially accurate and diverse items as negative samples to assist model training. GSN effectively addresses the accuracy-diversity dilemma and achieves improved diversity while maintaining accuracy with the help of a selective sampler. Moreover, a hyper-parameter in GSN allows for adjustment of the accuracy-diversity ratio of recommendation lists to satisfy the diverse demands. Compared to the state-of-the-art model, GSN improved R @20 by 1.62%, N @20 by 0.67%, G @20 by 3.59%, and E @20 by 4.15% on average over three real-world datasets, verifying the effectiveness of our proposed model in diversifying overall collaborative recommendations.

Disentangled Cascaded Graph Convolution Networks for Multi-Behavior Recommendation

Multi-behavioral recommender systems have emerged as a solution to address data sparsity and cold-start issues by incorporating auxiliary behaviors alongside target behaviors. However, existing models struggle to accurately capture varying user preferences across different behaviors and fail to account for diverse item preferences within behaviors. Various user preference factors (such as price or quality) entangled in the behavior may lead to sub-optimization problems. Furthermore, these models overlook the personalized nature of user behavioral preferences by employing uniform transformation networks for all users and items. To tackle these challenges, we propose the Disentangled Cascaded Graph Convolutional Network (Disen-CGCN), a novel multi-behavior recommendation model. Disen-CGCN employs disentangled representation techniques to effectively separate factors within user and item representations, ensuring their independence. In addition, it incorporates a multi-behavioral meta-network, enabling personalized feature transformation across user and item behaviors. Furthermore, an attention mechanism captures user preferences for different item factors within each behavior. By leveraging attention weights, we aggregate user and item embeddings separately for each behavior, computing preference scores that predict overall user preferences for items. Our evaluation of benchmark datasets demonstrates the superiority of Disen-CGCN over state-of-the-art models, showcasing an average performance improvement of 7.07% and 9.00% on respective datasets. These results highlight Disen-CGCN’s ability to effectively leverage multi-behavioral data, leading to more accurate recommendations.

Robotic Arm for Segregation Using Image Processing

The recent surge in industrial expansion can be attributed significantly to the advancements in automation technology. For better performance of industrial processes, automated systems are used. Image processing has played a great role in the applications of robotics and embedded systems. Sorting of objects is usually done by humans which takes a lot of time and effort. By employing image processing methods alongside appropriate sensors, object detection becomes feasible, enabling the utilization of robotic arms for sorting diverse items. This reduces human effort and improves the time to market the product. The model includes the conveyor system, capturing and detecting the objects and placing the objects. Using image processing techniques, the captured image is compared with its texture and color. Based on image processing, the robotic arm will be controlled and place the objects in desired locations. An Arduino model is used to process the captured image. The captured image is other than a pre-specified object, it involves the conveyor belt starting to move and the object is collected at the end of the conveyor system.

Choosing Among Undesirable Options: Children Consider Desirability of Available Choices in Evaluation of Socially Mindful Actions.

Previous studies show that adults and children evaluate the act of leaving a choice for others as prosocial, and have termed such actions as socially mindful actions. The current study investigates how the desirability of the available options (i.e., whether the available options are desirable or not) may influence adults' and children's evaluation of socially mindful actions. Children (N=120, 4- to 6-year-olds) and adults (N=124) were asked to evaluate characters selecting items for themselves from a set of three items-two identical items and one unique item-in a way that either leaves a choice (two diverse items) or leaves no choice (two identical items) for the next person (i.e., the beneficiary). We manipulated whether the available options were either desirable or undesirable (i.e., damaged). We found that adults' and 6-year-olds' evaluation of socially mindful actions is moderated by the desirability of the options. Although they evaluate the act of leaving a choice for others as nicer than the act of leaving no choice both when the choosing options are desirable and when they are undesirable, the discrepancy in the evaluation becomes significantly smaller when the choosing options are undesirable. We also found that inference of the beneficiary's feeling underlies social evaluation of the actor leaving a choice (or not). These findings suggest that children consider both the diversity of options left and the desirability of the available options in understanding and evaluating socially mindful acts.

Taking the Deer by the Antlers: Deer in Material Culture in the Balkan Neolithic

Prehistoric communities had strong ties with the animal world that surrounded them—animals were prey, sources of food, and raw materials, but also threats and mysteries, and certain animals often had an important place in the symbolic realm. With the process of domestication and the switch to animal husbandry as the main source of animal food, these relations changed considerably, and a certain dichotomy between “the domestic” and “the wild” may be noted in numerous past communities. When it comes to the Neolithic period in the Balkans, domestic animals had an important place in subsistence and economy, and it seems that cattle had a particularly prominent symbolic role. Wild species preserved some of their significance in both subsistence and symbolic realms, especially cervids (red deer, roe deer, and fallow deer). In this paper, the place of deer in the material culture of the Neolithic communities in the Balkans will be analysed: skeletal elements of deer were used for the production of diverse items, including non-utilitarian ones, or were part of ritual depositions, and deer representations are encountered in other materials, such as clay figurines. The symbolic meaning of deer cannot be reconstructed with certainty; however, it is probable that deer were tied with territoriality and the landscape.

The ubiquitous influence of image recognition in daily life

Image recognition technology has emerged due to the quick advancement of computer science. Using image recognition technology, a computer can identify and categorize diverse items in an image according to their functions. Artificial intelligence, image processing, and image recognition technology are all combined in this technology. The human eye and image recognition have a similar fundamental operating principle. The best outcomes can be achieved by taking out a few crucial elements from the image and contrasting them with the data that has been captured. This study describes how image recognition technology is used widely in modern life. It can improve people's lives if picture recognition is quick and precise. However, there are still many issues with picture recognition technology, such as security issues with gathering private information and loss of objectivity owing to algorithmic bias. Image recognition technology will become increasingly crucial in the connected world due to the ongoing development and enhancement of computer power. Image recognition technology will push what is technically possible in visual perception and artificial intelligence.

Graph Disentangled Contrastive Learning with Personalized Transfer for Cross-Domain Recommendation

Cross-Domain Recommendation (CDR) has been proven to effectively alleviate the data sparsity problem in Recommender System (RS). Recent CDR methods often disentangle user features into domain-invariant and domain-specific features for efficient cross-domain knowledge transfer. Despite showcasing robust performance, three crucial aspects remain unexplored for existing disentangled CDR approaches: i) The significance nuances of the interaction behaviors are ignored in generating disentangled features; ii) The user features are disentangled irrelevant to the individual items to be recommended; iii) The general knowledge transfer overlooks the user's personality when interacting with diverse items. To this end, we propose a Graph Disentangled Contrastive framework for CDR (GDCCDR) with personalized transfer by meta-networks. An adaptive parameter-free filter is proposed to gauge the significance of diverse interactions, thereby facilitating more refined disentangled representations. In sight of the success of Contrastive Learning (CL) in RS, we propose two CL-based constraints for item-aware disentanglement. Proximate CL ensures the coherence of domain-invariant features between domains, while eliminatory CL strives to disentangle features within each domains using mutual information between users and items. Finally, for domain-invariant features, we adopt meta-networks to achieve personalized transfer. Experimental results on four real-world datasets demonstrate the superiority of GDCCDR over state-of-the-art methods.

Investigating the diets and condition of Centrostephanus rodgersii (long-spined urchin) in barrens and macroalgae habitats in south-eastern Australia

Sea urchins are important herbivores that can graze macroalgae, creating ‘barren’ areas. These barrens are believed to offer little food for urchins due to the absence of attached macroalgae, leading to malnourishment as indicated by a low urchin gonad index. To investigate the diet and resultant gonad index of the long-spined sea urchin Centrostephanus rodgersii in New South Wales, Australia, we dissected 100 C. rodgersii individuals collected in macroalgae and barrens habitats along >470 km of SE NSW coastline. Diverse items were present in the digestive tracts of urchins from all habitats. These items included brown (42% barrens, 46% macroalgae), green (28% barrens, 42% macroalgae) and red algae (15% barrens, 12% macroalgae) and corallines (29% barrens, 37% macroalgae), molluscs (28% barrens, 29% macroalgae) and crustaceans (26% barrens, 22% macroalgae). There was no difference in urchin gut fullness between habitats (85% barrens, 90% macroalgae). Importantly, the gonad index only differed in macroalgae compared to barrens habitats at one location, with no differences detected at the other 4 locations. These results suggest that C. rodgersii has a diverse diet that is similar in both habitats, which could explain the similarities in gut fullness and gonad index. Our results suggest that C. rodgersii eat a broad diet including invertebrates and drift algae, and hence may not be malnourished in barrens. The finding of comparable gonad index between barrens and macroalgal areas further supports this conclusion. These findings challenge the prevailing perspective, indicating that sea urchins have sufficient food to survive and reproduce in different habitats.

KEGGSum: Summarizing Genomic Pathways

Over time, the renowned Kyoto Encyclopedia of Genes and Genomes (KEGG) has grown to become one of the most comprehensive online databases for biological procedures. The majority of the data are stored in the form of pathways, which are graphs that depict the relationships between the diverse items participating in biological procedures, such as genes and chemical compounds. However, the size, complexity, and diversity of these graphs make them difficult to explore and understand, as well as making it difficult to extract a clear conclusion regarding their most important components. In this regard, we present KEGGSum, a system enabling the efficient and effective summarization of KEGG pathways. KEGGSum receives a KEGG identifier (Kid) as an input, connects to the KEGG database, downloads a specialized form of the pathway, and determines the most important nodes in the graph. To identify the most important nodes in the KEGG graphs, we explore multiple centrality measures that have been proposed for generic graphs, showing their applicability to KEGG graphs as well. Then, we link the selected nodes in order to produce a summary graph out of the initial KEGG graph. Finally, our system visualizes the generated summary, enabling an understanding of the most important parts of the initial graph. We experimentally evaluate our system, and we show its advantages and benefits.

Promises and Pitfalls: Using Large Language Models to Generate Visualization Items.

Visualization items-factual questions about visualizations that ask viewers to accomplish visualization tasks-are regularly used in the field of information visualization as educational and evaluative materials. For example, researchers of visualization literacy require large, diverse banks of items to conduct studies where the same skill is measured repeatedly on the same participants. Yet, generating a large number of high-quality, diverse items requires significant time and expertise. To address the critical need for a large number of diverse visualization items in education and research, this paper investigates the potential for large language models (LLMs) to automate the generation of multiple-choice visualization items. Through an iterative design process, we develop the VILA (Visualization Items Generated by Large LAnguage Models) pipeline, for efficiently generating visualization items that measure people's ability to accomplish visualization tasks. We use the VILA pipeline to generate 1,404 candidate items across 12 chart types and 13 visualization tasks. In collaboration with 11 visualization experts, we develop an evaluation rulebook which we then use to rate the quality of all candidate items. The result is the VILA bank of ∼1,100 items. From this evaluation, we also identify and classify current limitations of the VILA pipeline, and discuss the role of human oversight in ensuring quality. In addition, we demonstrate an application of our work by creating a visualization literacy test, VILA-VLAT, which measures people's ability to complete a diverse set of tasks on various types of visualizations; comparing it to the existing VLAT, VILA-VLAT shows moderate to high convergent validity (R = 0.70). Lastly, we discuss the application areas of the VILA pipeline and the VILA bank and provide practical recommendations for their use. All supplemental materials are available at https://osf.io/ysrhq/.

Predicting diversification scores of videos in recommendation network

Recommendation systems are crucial in various platforms from shopping to online social networks. A key challenge in recommendation systems is to leverage diversity, exposing or recommending diverse items to individuals. Despite much effort on studying diversity in the recommendation systems, little work has focused on estimating how much an item will potentially affect user’s diversity experiences by contributing to consecutive recommendations in a session. In this work, we propose a deep learning model that can predict diversification scores, which is a degree of potential contribution to users’ diversity experiences of an item. The proposed model adopts multiple graph neural network layers with a novel attention mechanism that can capture the features of a given item and its related items in terms of recommendation. To prove the effectiveness of our approach, we collect a large dataset of video recommendations from YouTube and conduct random-walk experiments to simulate user traces. The evaluation results on the dataset shows that the proposed model accurately predicts each item’s contribution on user diversity experiences.

Optimizing Space Utilization In Container Packing

Container packing presents a complex optimization problem that seeks to efficiently pack diverse items into a fixed-size container. This study provides a comparative analysis of four algorithms -- Greedy Shelf, Shelfing with Rotation, Shelfing with Search, and Guillotine Paper Cutting -- investigating their proficiency in solving the container packing problem. Utilizing a set of four packages with differing dimensions and IDs, The study evaluated the performance of each algorithm. The results demonstrated that the Shelfing with Search algorithm outperformed its counterparts by yielding a stack height of 3019610 L units. Conversely, the Guillotine Paper Cutting algorithm performed poorly, with a stack height of 7537295 units. This research also explored the impact of different sorting methods on packing efficiency, revealing that sorting packages in descending order of height yields superior results. Consequently, this study provides an extensive evaluation of the various algorithms used for container packing, while suggesting promising directions for future research to enhance packing efficiency.

Deriving the number of salience maps an observer has from the number and quality of concurrent centroid judgments

[C. Koch, S. Ullman, Hum. Neurobiol.4, 219-227 (1985)] proposed a 2D topographical salience map that took feature-map outputs as its input and represented the importance "saliency" of the feature inputs at each location as a real number. The computation on the map, "winner-take-all," was used to predict action priority. We propose that the same or a similar map is used to compute centroid judgments, the center of a cloud of diverse items. [P. Sun, V. Chu, G. Sperling, Atten. Percept. Psychophys.83, 934-955 (2021)] demonstrated that following a 250-msec exposure of a 24-dot array of 3 intermixed colors, subjects could accurately report the centroid of each dot color, thereby indicating that these subjects had at least three salience maps. Here, we use a postcue, partial-report paradigm to determine how many more salience maps subjects might have. In 11 experiments, subjects viewed 0.3-s flashes of 28 to 32 item arrays composed of M, M = 3,...,8, different features followed by a cue to mouse-click the centroid of items of just the post-cued feature. Ideal detector response analyses show that subjects utilized at least 12 to 17 stimulus items. By determining whether a subject's performance in (M-1)-feature experiments could/could-not predict performance in M-feature experiments, we conclude that one subject has at least 7 and the other two have at least five salience maps. A computational model shows that the primary performance-limiting factors are channel capacity for representing so many concurrently presented groups of items and working-memory capacity for so many computed centroids.

A Novel Recommendation Algorithm Based on Multi-objective Evolution

During the past two decades, recommender systems have been playing an increasingly significant role in business decisions for providing personalized recommendations. However, the majority of the present algorithms attach excess importance to accuracy while overlooking users demand for diverse items. We have proposed a novel multi-objective evolutionary recommendation algorithm, which has the capability to look for compromise among accuracy and diversity. Moreover, due to the fact that Pareto optimal solutions of recommendation tend to be sparse, the proposed algorithm implements a new strategy of population initialization with the consideration of the sparse nature of the Pareto solutions. The proposed algorithm will take effect in the recommender systems.

DTGCF: Diversified Tag-Aware Recommendation with Graph Collaborative Filtering

In tag-aware recommender systems, users are strongly encouraged to utilize arbitrary tags to mark items of interest. These user-defined tags can be viewed as a bridge linking users and items. Most tag-aware recommendation models focus on improving the accuracy by introducing ingenious design or complicated structures to handle the tagging information appropriately. Beyond accuracy, diversity is considered to be another important indicator affecting the user satisfaction. Recommending more diverse items will provide more interesting items and commercial sales. Therefore, we propose a diversified tag-aware recommendation model based on graph collaborative filtering. The proposed model establishes a generic graph collaborative filtering framework tailored for tag-aware recommendations. To promote diversity, we adopt two modules: personalized category-boosted negative sampling to select a certain proportion of similar but negative items as negative samples for training, and adversarial learning to make the learned item representation category-free. To improve accuracy, we conduct a two-way TransTag regularization to model the relationship among users, items, and tags. Blending these modules into the proposed framework, we can optimize both the accuracy and diversity concurrently in an end-to-end manner. Experiments on Movielens datasets show that the proposed model can provide diverse recommendations while maintaining a high level of accuracy.

Influence of Educational Specialty on Perceptions towards Corporal Expression of Prospective Teachers.

Most educators overlook the importance of corporal expression, even though it has been shown to have various advantages for children of all ages. In the teaching-learning process, teacher views and beliefs are crucial because they have a significant impact on students. Therefore, the purpose of this research is to analyze the existing differences in the perceptions of future teachers towards corporal expression according to their gender and educational specialty. A total of 437 aspiring Spanish instructors participated in the sample, selected by the convenience sampling method, and answered the Questionnaire to Assess Perceptions of Corporal Expression in Future Spanish Teachers to gauge their understanding of and preparation for corporal expression via Google Forms. The Mann-Whitney U test was employed to evaluate the possible differences between the diverse items and factors according to gender and educational specialty. The results displayed good perceptions of corporal expression throughout the sample, finding significant differences in most items and all the dimensions when education specialty is considered. Nevertheless, gender variables did not seem to mediate those perceptions. Therefore, university degrees oriented to education must include the same amount of content related to corporal expression to ensure adequate initial training regardless of the stage of education at which the teachers develop their academic activity.

일본 수출규제 이후 자동차 산업의 무역경쟁력 변화에 관한 연구

Purpose This study analyzed how Japan’s export regulations affected the imports and exports of the automobile industry, and examined whether there was a change in the trade competitiveness of the automobile industry. Design/Methodology/Approach In order to examine the degree of improvement in the automobile industry’s imports, exports, and trade competitiveness with Japan in Korea, the world and Japanese import and export statistical data of the automobile industry were aggregated for five periods. In addition, the change trend in the dependence on imports and exports of automobiles and parts to Japan was examined. In addition, the change in competitiveness of the Korean automobile and parts industry to Japan was examined using the trade specialization index. Findings As a result of the analysis, it was found that the trade deficit of automobiles and parts with Japan in 2021 fell compared to 2017 and 2019. In the case of import and export dependence, there was little change in export dependence, but it was confirmed that there was a significant decrease in import dependence. Korea’s dependence on exports of automobiles and parts to Japan has decreased slightly, but its dependence on imports has decreased by 10 times, which can be interpreted as Japan receiving more trade losses in the automobile industry due to the Korea-Japan trade dispute. Research Implications In the passenger car sector, it can be evaluated that there has been a result of the boycott of Japanese products, and efforts to strengthen the competitiveness of domestic cars and diversify import lines should continue to continue this in the future. In addition, in order to continue this performance in automobile parts, efforts should be made to improve product competitiveness and import dependence in more diverse item groups in the parts.

Context Uncertainty in Contextual Bandits with Applications to Recommender Systems

Recurrent neural networks have proven effective in modeling sequential user feedbacks for recommender systems. However, they usually focus solely on item relevance and fail to effectively explore diverse items for users, therefore harming the system performance in the long run. To address this problem, we propose a new type of recurrent neural networks, dubbed recurrent exploration networks (REN), to jointly perform representation learning and effective exploration in the latent space. REN tries to balance relevance and exploration while taking into account the uncertainty in the representations. Our theoretical analysis shows that REN can preserve the rate-optimal sublinear regret even when there exists uncertainty in the learned representations. Our empirical study demonstrates that REN can achieve satisfactory long-term rewards on both synthetic and real-world recommendation datasets, outperforming state-of-the-art models.

Compositional tourists’ expenditure: Modeling through Dirichlet regression

This study models the relative expenditure of tourists in terms of budget allocation according to their dependence on a group of covariates. A model that captures how tourist distributes its budget among the diverse items is introduced to characterize and compare different types of tourists according to their relative expenditure patterns. For the empirical exercise, data for the period 2014–2019 collected by the Ministry of Tourism from the Inbound Tourism survey in Uruguay is analyzed by means of the compositional data analysis and modeled by a Dirichlet regression. The empirical results show that the expending pattern in accommodation, food, and others items depend on the destination, the season, the nationality, and the type of accommodation. In addition, the inferential analysis reveals different typologies of tourist, providing a novel interpretation of the tourist behavior from the microeconomic perspective. JEL ClassificationL83, Z30, C14.

To Recommend or Not? A Model-Based Comparison of Item-Matching Processes

Recommender systems are central to modern online platforms, but a popular concern is that they may be pulling society in dangerous directions (e.g., towards filter bubbles). However, a challenge with measuring the effects of recommender systems is how to compare user outcomes under these systems to outcomes under a credible counterfactual world without such systems. We take a model-based approach to this challenge, introducing a dichotomy of process models that we can compare: (1) a “recommender” model describing a generic item-matching process under a personalized recommender system and (2) an “organic” model describing a baseline counterfactual where users search for items without the mediation of any system. Our key finding is that the recommender and organic models result in dramatically different outcomes at both the individual and societal level, as supported by theorems and simulation experiments with real data. The two process models also induce different trade-offs: regularization improves the mean squared error of matches in both settings, but at the cost of less diverse (less radical) items chosen in the recommender model but more diverse (more radical) items chosen in the organic model. These findings provide a formal framework for how recommender systems may be fundamentally altering how we interact with content, in a world increasingly mediated by such systems.