Item Features Research Articles

Model-driven approach running route two-level SVD with context information and feature entities in recommender system

In digital businesses, the offers of goods and services to users in recommendation systems are generally based on the features of the items and the demands of the users. Recommendation systems use contextual information, such as location, time, activity, individual, and relationship to generate relevant and even personalized recommendations. A large proportion of users do not always rate items that they have used which creates the challenge of sparse data in recommendation systems. On the other hand, when a new user or item enters the system, it does not have much information about it. This leads to the challenge a cold start. In this paper, a new hybrid method is presented to reduce the challenge of cold start and sparse data based on the items desired by users according to the similarity measure of user-contextual information, item-contextual information, and two-level singular value decomposition. Two data sets IMDB and STS due to the exerting of user's feature, items feature, and contextual information to review the proposed method. In order to the accuracy of the prediction of the criterion MAE and RMSE with an accuracy of 95%. However, since the user's rating of the item is of particular importance in the recommender systems. The method CSSVD compared to the methods TF, CACF, CTLSVD, and MF-LOD is used evaluation measure Precision, Recall, F1-score. The results contribute to improving the accuracy of the suggestions given to users in recommender systems.

A two‐phase knowledge distillation model for graph convolutional network‐based recommendation

Graph convolutional network (GCN)-based recommendation has recently attracted significant attention in the recommender system community. Although current studies propose various GCNs to improve recommendation performance, existing methods suffer from two main limitations. First, user–item interaction data is generally sparse in practice, highlighting these methods' ineffectiveness in learning user and item feature representations. Second, they usually perform a dot-product operation to model and calculate user preferences on items, leading to inaccurate user preference learning. To address these limitations, this study adopts a design idea that sharply differs from existing works. Specifically, we introduce the knowledge distillation concept into GCN-based recommendation and propose a two-phase knowledge distillation model (TKDM) improving recommendation performance. In Phase I, a self-distillation method on a graph auto-encoder learns the user and item feature representations. This auto-encoder employs a simple two-layer GCN as an encoder and a fully connected layer as a decoder. On this basis, in Phase II, a mutual-distillation method on a fully connected layer is introduced to learn user preferences on items with triple-based Bayesian personalized ranking. Extensive experiments on three real-world data sets demonstrate that TKDM outperforms classic and state-of-the-art methods related to GCN-based recommendation problems.

Weighted AutoEncoding recommender system

AbstractRecommender systems are information filtering tools that seek to match customers with products or services of interest. Most of the prevalent collaborative filtering recommender systems, such as matrix factorization and AutoRec, suffer from the “cold‐start” problem, where they fail to provide meaningful recommendations for new users or new items due to informative‐missing from the training data. To address this problem, we propose a weighted AutoEncoding model to leverage information from other users or items that share similar characteristics. The proposed method provides an effective strategy for borrowing strength from user or item‐specific clustering structure as well as pairwise similarity in the training data, while achieving high computational efficiency and dimension reduction, and preserving nonlinear relationships between user preferences and item features. Simulation studies and applications to three real datasets show advantages in prediction accuracy of the proposed model compared to current state‐of‐the‐art approaches.

Dual Auto-Encoder Based Rating Prediction Recommendation Algorithm

Collaborative filtering is the most widely used method in recommendation algorithms, but it still faces the serious problem of data sparsity. Traditional collaborative filtering uses matrix decomposition to learn the latent features of users and items. As an extension model of matrix decomposition, Funk-SVD model has attracted wide attention due to its good scalability and easy implementation, but it is difficult to extract the latent features of users and items from sparse rating information because it essentially learns the linear relationship between users and items. To solve this problem, we propose a Dual auto-encoder based Rating Prediction Recommendation Algorithm (DRPRA) model. The DRPRA model uses the strong ability of deep learning in feature learning, which combines double auto-encoders with Funk-SVD. First, the auto-encoder captures the latent features of users and items respectively. Then, the Funk-SVD combines the user features with item features to reconstruct the rating matrix. After that, we minimize the error between original rating matrix and reconstructed rating matrix, and to alleviate the problem of data sparsity and improve the accuracy of rating prediction effectively. We conducted extensive experiments on Movielens-100K, Movie Tweeting-10k, and Film Trust datasets, and the results show that the rating prediction model based on dual auto-encoders has a superior recommendation performance.

Refining Graph Representation for Cross-Domain Recommendation Based on Edge Pruning in Latent Space

In this paper, we present refining graph representation for cross-domain recommendation (CDR) based on edge pruning considering feature distribution in a latent space. Conventional graph-based CDR methods have utilized all ratings and purchase histories of user’s products. However, some items purchased by users are not related to the domain for recommendation, and this information becomes noise when making CDR. So, the proposed method introduces edge pruning into the latest graph-based CDR method to refine graph representation. To compare the item embedding features calculated in different domains, we construct a latent space and perform edge pruning through their correlations. Additionally, we introduce a state-of-the-art graph neural network into the graph construction of the proposed method that considers the interactions between users and items thereby obtaining effective embedding features in a domain. This makes it possible to consider domain-specific user preferences and estimate embedding features with high-expressive power. Furthermore, to compare the embedding features of items in the two domains, we construct their latent spaces and project them. Edge pruning is performed using the correlation of items between the two domains on the latent space. We obtain cross domain specific graph representation through edge pruning, which improves the performance by considering the relationship between both items across domains. To the best of our knowledge, no study in the CDR field focuses on eliminating unnecessary node information. We have demonstrated the effectiveness of the proposed method by comparing several graph-based state-of-the-art methods.

Hierarchical User Intent Graph Network for Multimedia Recommendation

Understanding user preference on item context is the key to acquire a high-quality multimedia recommendation. Typically, the pre-existing features of items are derived from pre-trained models (e.g. visual features of micro-videos extracted from some neural networks), and then introduced into the recommendation framework (e.g. collaborative filtering) to capture user preference. However, we argue that such a paradigm is insufficient to output satisfactory user representations, which hardly profile personal interests well. The key reason is that present works largely leave user intents untouched, then failing to encode such informative representation of users. In this work, we aim to learn multi-level user intents from the co-interacted patterns of items, so as to obtain high-quality representations of users and items and further enhance the recommendation performance. Towards this end, we develop a novel framework, <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">Hierarchical User Intent Graph Network</i> , which exhibits user intents in a hierarchical graph structure, from the fine-grained to coarse-grained intents. In particular, we get the multi-level user intents by recursively performing two operations: 1) intra-level aggregation, which distills the signal pertinent to user intents from co-interacted item graphs; and 2) inter-level aggregation, which constitutes the supernode in higher levels to model coarser-grained user intents via gathering the nodes’ representations in the lower ones. Then, we refine the user and item representations as a distribution over the discovered intents, instead of simple pre-existing features. To demonstrate the effectiveness of our model, we conducted extensive experiments on three public datasets. Our model achieves significant improvements over the state-of-the-art methods, including MMGCN and DisenGCN. Furthermore, by visualizing the item representations, we provide the semantics of user intents.

Multi-Modal Variational Graph Auto-Encoder for Recommendation Systems

Graph embedding based methods have been used in recommendation systems recently, owing to their advances in modeling nodes as embeddings in a low-dimensional space. By effective neighborhood aggregation, graph convolutional networks can exploit high-order connections of neighbors such that the learned embeddings could be more informative thus improve the recommendation performance. However, user and item representations learned by graph aggregation inherently contain uncertainty due to sparsity of user-item interactions and noise of item features. To address these challenges, we propose a multi-modal variational graph auto-encoder (MVGAE) method. Specifically, we design modality-specific variational encoders that learn a Gaussian variable for each node whereas the mean vector represents semantic information and the variance vector denotes the noise level of the corresponding modality. Moreover, with the conditional independence assumption, the modality-specific Gaussian node embeddings are fused according to the product-of-experts principle, where the semantic information in each modality is weighted based on the estimated uncertainty level. Extensive experiments on three public datasets, Amazon Movies, Amazon Electronics and AliShop-7 C, demonstrate that our proposed method achieves competitive performance when compared with the state-of-the-art algorithms.

An Web Service Recommendation Method Based on Location and Decomposition Machine Model

With the increasing number of Web services with similar functions on the Internet, traditional collaborative filtering service recommendation methods may encounter problems such as data sparseness, cold start, and poor scalability. To solve the above problems, this project proposes a new Web service recommendation method based on the decomposition machine model. The method decomposes the user trust relationship matrix and the product rating matrix while adding the geographic location information of the service, and transforms the correlation matrix of the calculated user feature vector and item feature vector into the same latent factor space by means of a decomposition machine. Optimize training model parameters to provide users with accurate prediction scores. The ultimate purpose of QoS prediction is to recommend high-quality services to users, improve the efficiency of users' discovery and selection of high-quality services, and ultimately promote the utilization of network Web services and promote service providers to release higher-quality services. Scientific significance, but also has better application value.

Enhancing rating prediction for recommendation by metric learning with gray relational analysis

PurposeThe purpose of the paper is to improve the rating prediction accuracy in recommender systems (RSs) by metric learning (ML) method. The similarity metric of user and item is calculated with gray relational analysis.Design/methodology/approachFirst, the potential features of users and items are captured by exploiting ML, such that the rating prediction can be performed. In metric space, the user and item positions can be learned by training their embedding vectors. Second, instead of the traditional distance measurements, the gray relational analysis is employed in the evaluation of the position similarity between user and item, because the latter can reduce the impact of data sparsity and further explore the rating data correlation. On the basis of the above improvements, a new rating prediction algorithm is proposed. Experiments are implemented to validate the effectiveness of the algorithm.FindingsThe novel algorithm is evaluated by the extensive experiments on two real-world datasets. Experimental results demonstrate that the proposed model achieves remarkable performance on the rating prediction task.Practical implicationsThe rating prediction algorithm is adopted to predict the users' preference, and then, it provides personalized recommendations for users. In fact, this method can expand to the field of classification and provide potentials for this domain.Originality/valueThe algorithm can uncover the finer grained preference by ML. Furthermore, the similarity can be measured using gray relational analysis, which can mitigate the limitation of data sparsity.

Ice Is Hot and Water Is Dry

Abstract: To monitor students’ progress and adapt instruction to students’ needs, teachers increasingly use repeated assessments of equivalent tests. The present study investigates whether equivalent reading tests can be successfully developed via rule-based item design. Based on theoretical considerations, we identified 3-item features for reading comprehension at the word, sentence, and text levels, respectively, which should influence the difficulty and time intensity of reading processes. Using optimal design algorithms, a design matrix was calculated, and four equivalent test forms of the German reading test series for second graders (quop-L2) were developed. A total of N = 7,751 students completed the tests. We estimated item difficulty and time intensity parameters as well as person ability and speed parameters using bivariate item response theory (IRT) models, and we investigated the influence of item features on item parameters. Results indicate that all item properties significantly affected either item difficulty or response time. Moreover, as indicated by the IRT-based test information functions and analyses of variance, the four different test forms showed similar levels of difficulty and time-intensity at the word, sentence, and text levels (all η2&lt; .002). Results were successfully cross-validated using a sample of N = 5,654 students.

Towards optimizing the measurement of self‐perceived cognitive functioning: Examination of harmonized subjective cognitive items

AbstractBackgroundFor the item analysis project of the Subjective Cognitive Decline Initiative (SCD‐I) Working Group, we have pooled item level data from North American, European, and Australian cognitive aging studies. As a step toward developing a questionnaire that can be used across settings and studies, we examined similarly worded groups of items and compared their psychometric quality.MethodItem‐level subjective cognitive data were combined from 21 studies (&gt;18,000 participants). Unidimensional models consistent with item response theory captured a general subjective cognition factor using ADNI as the reference sample. Other studies were calibrated by using parameter constraints on common items. We characterized the measurement properties of 648 items from 35 measures utilized in the contributing studies. Next, we identified 20 groups of items (median items per group=6) with overlapping content from our previous study (Rabin et al., 2015), and plotted their item information curves, reflecting the items’ ability to discriminate between trait scores at different levels of the subjective cognition factor score. We then considered items with the highest peak information across groupings and identified commonalities in key features such as terminology of item stems, temporal referents, and response options.ResultWe present instrument characteristics (Table 1) and item information curves for 6 groupings (Figure 1). For items with the greatest information, we observed trends for the following features: clarity and directness of language of stem, utilization of current as opposed to other time referents, assessment of ability/level of difficulty rather than change over time or frequency of problem (Table 2).ConclusionWe identified key linguistic/structural features of subjective cognitive functioning items with the greatest contributions to measurement precision. We next plan to analyze all item groupings in our test bank to facilitate construction of a measure with high quality items from cognitive domains of interest (e.g., memory, language, executive functioning). Ultimately, we hope to derive a subjective cognitive screening tool that enables reliable identification of subtle cognitive changes in clinically normal individuals.

Modeling sentimental bias and temporal dynamics for adaptive deep recommendation system

Recommendation systems rely on the historic data of users' purchases and their feedbacks to profile their preferences and make future recommendations. Most of these systems usually employ Collaborative Filtering (CF) models to analyze users’ ratings and infer the latent factors which represent the user and item features in k-dimensional latent space. However, the historical rating data used for recommendations are usually sparsed and unbalanced. Various approaches have been used to resolve these issues by combining the user’s ratings and reviews to better capture the user’s sentiments and make accurate recommendations. Other challenges comprise changes in users’ preferences and items’ perceptions over time. Therefore, this paper presents a new Sentiment Scoring Model (SSM) based on Long-/Short-Term Memory and a combination function that catches the sentiment bias between user rating and review to relieve the sparsity and unbalanced dataset. Next, we proposed an Adaptive LSTM (ALSTM) method that can model the drifting of user and item features to improve the recommendation accuracy. We show the performance of our model on the three real-world rating datasets from Amazon reviews, which comprises Fine Food, Baby, and Cell-phone & Accessories categories. The result shows the superiority of our proposed model over the existing static and dynamic models. The statistical test shows that all the performance gains are significant at p < 0.05.

Context Matters: Accounting for Item Features in the Assessment of Citizen Scientists’ Scientific Reasoning Skills

Citizen science (CS) projects engage citizens for research purposes and promote individual learning outcomes such as scientific reasoning (SR) skills. SR refers to participants’ skills to solve problems scientifically. However, the evaluation of CS projects’ effects on learning outcomes has suffered from a lack of assessment instruments and resources. Assessments of SR have most often been validated in the context of formal education. They do not contextualize items to be authentic or to represent a wide variety of disciplines and contexts in CS research. Here, we describe the development of an assessment instrument that can be flexibly adapted to different CS research contexts. Furthermore, we show that this assessment instrument, the SR questionnaire, provides valid conclusions about participants’ SR skills. We found that the deep-structure and surface features of the items in the SR questionnaire represent the thinking processes associated with SR to a substantial extent. We suggest that practitioners and researchers consider these item features in future adaptations of the SR questionnaire. This will most likely enable them to draw valid conclusions about participants’ SR skills and to gain a deeper understanding of participants’ SR skills in CS project evaluation.

Mechanisms of feature binding in visual working memory are stable over long delays.

The ability to accurately retain the binding between the features of different objects is a critical element of visual working memory. The underlying mechanism can be elucidated by analyzing correlations of response errors in dual-report experiments, in which participants have to report two features of a single item from a previously viewed stimulus array. Results from separate previous studies using different cueing conditions have indicated that location takes a privileged role in mediating binding between other features, in that largely independent response errors have been observed when location was used as a cue, but errors were highly correlated when location was one of the reported features. Earlier results from change detection tasks likewise support such a special role of location, but they also suggest that this role is substantially reduced for longer retention intervals in favor of object-based representation. In the present study, we replicated the findings of previous dual-report tasks with different cueing conditions, using matched stimuli and procedures. Moreover, we show that the observed patterns of error correlations remain qualitatively unchanged with longer retention intervals. Fits with neural population models demonstrate that the behavioral results at long, as well as short, delays are best explained by memory representations in independent feature maps, in which an item's features are bound to each other only via their shared location.

Hypergraph Neural Networks with Attention Mechanism for Session-based Recommendation

The session-based recommendation task is designed to predict the behavior of the current session at the next moment based on multiple anonymous sessions. Due to the lack of user information in the session, the traditional recommendation model cannot be used directly to model the interest of specific users. In this paper, a session recommendation model based on hypergraph neural networks and attention mechanism (HGNNA) is proposed. Firstly, the features of items are learned by constructing hypergraph neural networks, then the conversation information is aggregated by self-attention mechanism, and finally the information among similar sessions is aggregated by graph attention networks. The hypergraph neural networks can capture the correlation between items, the self-attention mechanism can show the interest of the current session, and the graph attention networks can find the interest pattern between similar sessions, so that the representation vector of the session includes the information of the items in the session, other items outside the session and other sessions. In the experiments on two datasets, Yoochoose1/4 and Diginetica, the recommendation effect of HGNNA is higher than that of other relevant methods, especially in the P@20, which is improved by 0.69 and 1.40 respectively.

Neural matrix factorization++ based recommendation system.

In recent years, Recommender System (RS) research work has covered a wide variety of Artificial Intelligence techniques, ranging from traditional Matrix Factorization (MF) to complex Deep Neural Networks (DNN). Traditional Collaborative Filtering (CF) recommendation methods such as MF, have limited learning capabilities as it only considers the linear combination between user and item vectors. For learning non-linear relationships, methods like Neural Collaborative Filtering (NCF) incorporate DNN into CF methods. Though, CF methods still suffer from cold start and data sparsity. This paper proposes an improved hybrid-based RS, namely Neural Matrix Factorization++ (NeuMF++), for effectively learning user and item features to improve recommendation accuracy and alleviate cold start and data sparsity. NeuMF++ is proposed by incorporating effective latent representation into NeuMF via Stacked Denoising Autoencoders (SDAE). NeuMF++ can also be seen as the fusion of GMF++ and MLP++. NeuMF is an NCF framework which associates with GMF (Generalized Matrix Factorization) and MLP (Multilayer Perceptrons). NeuMF achieves state-of-the-art results due to the integration of GMF linearity and MLP non-linearity. Concurrently, incorporating latent representations has shown tremendous improvement in GMF and MLP, which result in GMF++ and MLP++. Latent representation obtained through the SDAEs' latent space allows NeuMF++ to effectively learn user and item features, significantly enhancing its learning capability. However, sharing feature extractions among GMF++ and MLP++ in NeuMF++ might hinder its performance. Hence, allowing GMF++ and MLP++ to learn separate features provides more flexibility and greatly improves its performance. Experiments performed on a real-world dataset have demonstrated that NeuMF++ achieves an outstanding result of a test root-mean-square error of 0.8681. In future work, we can extend NeuMF++ by introducing other auxiliary information like text or images. Different neural network building blocks can also be integrated into NeuMF++ to form a more robust recommendation model.

Scattered accumulation hotspots of macro-litter on the seafloor: Insights for mitigation actions

Marine litter is an ever-increasing problem that demands immediate reduction plans and mitigation actions that should act synergically to efficiently meet ambitious goals. Since the seafloor has been recognized as the major sink for marine debris, the study of litter accumulation dynamics represent a fundamental tool to evaluate possible removal actions. We analysed a 7 years (2013–2019) standardized data series collected along Sardinian fishing grounds through MEDiterranean International Trawl Survey, for which estimates of density and weight of seafloor macro-litter were calculated for over 707 hauls. Results show the absence of any temporal trend in seafloor macro-litter density and weight, but rather indicate a spatial and bathymetric segregation of different litter categories. Our data showed how different sources and physical features of macro-litter items (i.e., plastic and fishing gear, rubber, glass, metals, and cloth) led to spatially segregated accumulation hotspots. We also point out here how the identification of seafloor macro-litter hotspots using aggregated data that include plastic items could obscure the identification of other segregated but yet relevant hotspots of other macro-litter categories accumulated in the marine environment. These hotspots often occurred at shallower depths and closer to coastlines, thus representing potential spots where eventual future litter removal action could be prioritized.

User-location distribution serves as a useful feature in item-based collaborative filtering

Personalized recommender system is a powerful method to solve the problem of information overload, which has been widely applied in a variety of scenarios, such as e-commerce, video platforms and social networks, to help users find relevant items or friends of interest. Collaborative filtering is the most successful and widely used algorithm in the recommender systems as its powerful capability of generating recommendations by sharing collective experiences of users. In recent years, the use of mobile devices and the rapid development of internet infrastructures provide the possibility to analyze regional features of items based on user locations. Here we improve the performance of collaborative filtering by using user-location distribution to uncover the potential similarities between items. We find that the similarity of user-location distribution is one efficient measure for the item–item similarities in the framework of collaborative filtering to generate personalized recommendation for users. Furthermore, we have also mixed similarity measures of user-location distribution and the traditional method based on the number of common users linearly to optimize the performance of collaborative filtering. Based on the Movielens data set, we show that the performance of our methods could be improved in terms of the metrics of accuracy and diversity simultaneously.

Chomsky’s London

Semantic externalism is the view according to which proper names and other nominals have the capacity to refer to language-independent objects. On this view, the proper name ‘London’ is related semantically to a worldly object, London. Chomsky’s long held position is that this relational conception of reference is untenable. According to his internalist framework, semantics should be restricted to the examination of the informational features of I-language items. Externalists reject this restriction by saying that without employing the relational notion of reference, it would remain entirely mysterious how we can talk about our perceptible environment. This paper offers a novel argument for externalism. The basic idea is that external reference proves to be indispensable even for Chomskyans who regard our talk about the environment as irrelevant for the purposes of semantics.

Recommendation System for Voters using Classifier Rule

Abstract: Recommender System gives suggestions based on the user’s preferences and features of items. Ultimately its performance and efficiency depends on these factors and their representations. It also reduces the uncertainties by displaying item features. Here fuzzy logic in the recommendation system plays vital role to handle uncertainty. Fuzzy logic helps the user to provide most proper information related to the model. The multiple opinions from the user are taken on the basis of Voters database to take appropriate decision. In the proposed system classification of the database is done with the help of rules of decision tree and PART in which all the attributes are included whichever given. Comparison between these proves that PART works more efficiently than that of rules of decision tree. Keywords: Recommendation system for voters, Rules of decision tree, PART