Historical Interaction Data Research Articles

Vector Storage Based Long-term Memory Research on LLM

Abstract Current large language model (LLM) intelligences face the challenges of high inference cost and low decision quality when dealing with complex tasks, and are especially deficient in maintaining context coherence during long tasks. This research presents an innovative vector storage long-term memory mechanism model (VIMBank) to enhance the long-term context retention ability and task execution efficiency of LLM intelligences by storing and retrieving historical interaction data through a vector database. VIMBank utilizes a dynamic memory updating strategy and the Ebbinghaus forgetting curve theory to efficiently manage the memory of intelligences and reinforce critical information, forgetting unimportant data, and optimizing storage and reasoning costs. The experimental results show that VIMBank significantly improves the decision quality and efficiency of LLM intelligences in multi-tasking scenarios and reduces the computational cost. Compared with different agents, the success rate of task decision is increased by 10% to 20%, and the reasoning cost is reduced by about 23%, which provides an important theoretical basis and practical support for the future development of intelligences with long term memory and adaptive learning ability.

Boosting Healthiness Exposure in Category-Constrained Meal Recommendation Using Nutritional Standards

Food computing, a newly emerging topic, is closely linked to human life through computational methodologies. Meal recommendation, a food-related study about human health, aims to provide users a meal with courses constrained from specific categories (e.g., appetizers, main dishes) that can be enjoyed as a service. Historical interaction data, important user information, is often used by existing models to learn user preferences. However, if a user’s preferences favor less healthy meals, the model will follow that preference and make similar recommendations, potentially negatively impacting the user’s long-term health. This emphasizes the necessity for health-oriented and responsible meal recommendation systems. In this article, we propose a healthiness-aware and category-wise meal recommendation model called CateRec, which boosts healthiness exposure by using nutritional standards as knowledge to guide the model training. Two fundamental questions are raised and answered: (1) How can the healthiness of meals be evaluated? Two well-known nutritional standards from the World Health Organization and the United Kingdom Food Standards Agency are used to calculate the healthiness score of the meal. (2) How can the model training be guided in a health-oriented manner? We construct category-wise personalization partial rankings and category-wise healthiness partial rankings, and theoretically analyze that they meet the necessary properties and assumptions required to be trained by the maximum posterior estimator under Bayesian probability. The data analysis confirms the existence of user preferences leaning towards less healthy meals in two public datasets. A comprehensive experiment demonstrates that our CateRec effectively boosts healthiness exposure in terms of mean healthiness score and ranking exposure while being comparable to the state-of-the-art model in terms of recommendation accuracy.

KGCF: Social relationship-aware graph collaborative filtering for recommendation

With the increasing popularity of recommendation techniques and social networks, social network recommendation has become a significant research field, i.e., predicting a user's preferences based on her or his historical interaction data, because the social relationships of users can not only enrich their interaction information, but also imply the accurate characterization of their preferences. Although there are some existing studies that predict users' preferences by using the characteristics of social relationships, their time complexity and hardware resource consumption are often very expensive, which limits the feasibility in large-scale real-world scenarios. To address these issues, in this paper, we propose a simple and effective K-core Graph Collaborative Filtering (KGCF) model to incorporate the user's social features into a recommendation framework. The user's interaction information is auto-encoded linearly by incorporating the social relationship graph into the user's interaction information and constructing a multi-layer graph filter. The linear autoencoder graph filter alleviates the highly sparse data problem and dramatically reduces the training time and hardware resource consumption. Experimental results on several real-world datasets show the superior effectiveness and spatiotemporal efficiency compared with the existing baselines, especially the training speed is significantly accelerated for large-scale datasets.

A Collaborative Filtering-Based Recommender Systems approach for Multifarious applications

Recommender systems are crucial in today's IT landscape, enhancing user experiences in various industries. Collaborative filtering (CF) is a key approach, using historical interaction data to predict user preferences. This paper presents CF advancements, with a focus on latent factor models that represent users and items in a compact feature space. It also addresses sparsity issues with techniques like neighborhood-based approaches and content augmentation. Contextual CF, which incorporates temporal and contextual dynamics, is explored through methods like matrix factorization with side information and session-based recommendation. Evaluation metrics such as MAE and RMSE, along with novel ranking-based metrics, provide a comprehensive assessment of recommendation quality. In this paper we outline cutting-edge CF techniques, showcasing their mechanisms and applications and were able to achieve accurate recommendations of almost 90% using MAE and RMSE metrics. By integrating latent factor modeling, sparsity mitigation, contextual enrichment, and advanced evaluation, it paves the way for the next generation of personalized recommendation systems, tailored to meet evolving demands in modern information environments.

MMKG-PAR: Multi-Modal Knowledge Graphs-Based Personalized Attraction Recommendation

As the tourism industry rapidly develops, providing personalized attraction recommendations has become a hot research area. Knowledge graphs, with their rich semantic information and entity relationships, not only enhance the accuracy and personalization of recommendation systems but also energize the sustainable development of the tourism industry. Current research mainly focuses on single-modal knowledge modeling, limiting the in-depth understanding of complex entity characteristics and relationships. To address this challenge, this paper proposes a multi-modal knowledge graphs-based personalized attraction recommendation (MMKG-PAR) model. We utilized data from the “Travel Yunnan” app, along with users’ historical interaction data, to construct a collaborative multi-modal knowledge graph for Yunnan tourist attractions, which includes various forms such as images and text. Then, we employed advanced feature extraction methods to extract useful features from multi-modal data (images and text), and these were used as entity attributes to enhance the representation of entity nodes. To more effectively process graph-structured data and capture the complex relationships between nodes, our model incorporated graph neural networks and introduced an attention mechanism for mining and inferring higher-order information about entities. Additionally, MMKG-PAR introduced a dynamic time-weighted strategy for representing users, effectively capturing and precisely describing the dynamics of user behavior. Experimental results demonstrate that MMKG-PAR surpasses existing methods in personalized recommendations, providing significant support for the continuous development and innovation in the tourism industry.

AdaMO: Adaptive Meta-Optimization for cold-start recommendation

Meta-learning has been proven effective in tackling the cold-start problem of recommendation systems. Most work in this line adopts the meta-optimization idea that learns global knowledge to initialize the base recommender and adapts it for unseen recommendation tasks. Typically, the process is non-adaptive in the sense that both the initialization and model fitting are not task-specific. Such non-adaptive approaches do not work well for the strict cold-start problem, where historical interaction data of the new users is scarce or not available at all. In this paper, we propose a novel adaptive meta-optimization approach, called Adaptive Meta-Optimization (AdaMO), to address the problem. AdaMO is fully adaptive, namely, it simultaneously customizes the initialization and model fitting according to task-specific information. During the cold-start phase, it aggregates task-specific information with transferred knowledge from all historical tasks to tune the initialization of the base model. In the subsequent warm-up phase, AdaMO mines the adaptive transition information from the behavior patterns and parameter status to effectively fit the model using transitional learning. By making a sophisticated balance between shared knowledge and task-specific information, AdaMO outperforms the state-of-the-art methods in various cold-start scenarios.

BCE4ZSR: Bi-encoder empowered by teacher cross-encoder for zero-shot cold-start news recommendation

In the realm of news recommendations, the persistent challenge of the cold-start problem continues to impede progress. Existing approaches rely heavily on information exchange between news articles and users to personalize news recommendations and have struggled to adapt to users and news articles without historical interaction data. In this paper, we introduce BCE4ZSR, a novel framework that leverages a rarely utilized zero-shot approach to effectively tackle the cold-start problem in news recommendations. The proposed approach consists of two main steps: First, we generate embeddings for inference with a sentence transformer (a bi-encoder). In the second step, a fine-tuned transformer model is augmented during the training phase to distil the bi-encoder with knowledge from the cross-encoder model using a student–teacher framework. As a result of this synergy, the cross-encoder serves as a teacher, imparting its knowledge to the bi-encoder. The proposed technique can be applied to any neural news recommender system and is empirically evaluated in both cold-start and regular user-news interaction situations. Experiments on real-world benchmark datasets (MIND-small and MIND-large) indicated that BCE4ZSR outperformed the baseline methods in terms of nDCG@k, AUC, and MRR. Specifically, AUC improvement of 1.5%–6% & 2.2%–7.93%; MRR improved by 1%–13.85% & 0.9%–14.72%; nDCG@5 improved by 4.9%–18.79% & 2.1%–16.44%; and nDCG@10 improved by 1.7%–15.18% & 1.9%–14.24% in the cold start and warm user scenarios respectively, proving the superiority of our model compared to baseline methods.

ASKAT: Aspect Sentiment Knowledge Graph Attention Network for Recommendation

In modern online life, recommender systems can help us filter unimportant information. Researchers of recommendation algorithms usually utilize historical interaction data to mine potential user preferences. However, most existing methods use rating data to mine user interest preferences, ignoring rich textual information such as reviews. Although some researchers have attempted to combine ratings and reviews for recommendation, we believe the following shortcomings still exist. First, existing methods are overly dependent on the accuracy of external sentiment analysis tools. Second, existing methods do not fully utilize the features extracted from reviews. Further, existing methods focus only on the aspects that users like, while ignoring the aspects that users dislike, and they cannot completely model users’ true preferences. To address the above issues, in this paper, we propose a recommendation model based on the aspect of the sentiment knowledge graph attention network (ASKAT). We first use the improved aspect-based sentiment analysis algorithm to extract aspectual sentiment features from reviews. Then, to overcome the difficulty in underutilizing the information extracted from the comments, we build aspects of sentiment-enhanced collaborative knowledge mapping. After that, we propose a new graph attention network that uses sentiment-aware attention mechanisms to aggregate neighbour information. Finally, our experimental results on three datasets, Movie, Amazon book, and Yelp, show that our model consistently outperforms the baseline model in two recommendation scenarios, click-through-rate prediction and Top-k recommendation. Compared with other models, the method shows significant improvement in both recommendation accuracy and personalised recommendation effectiveness.

Doubly constrained offline reinforcement learning for learning path recommendation

Learning path recommendation refers to the application of interactive recommendation systems in the field of education, aimed at optimizing learning outcomes while minimizing the workload of learners, teachers, and curriculum designers. Reinforcement Learning (RL) has proven effective in capturing and modeling the complex interactions among course activities, learner behaviors, and educational outcomes. Therefore, combining the two approaches presents endless possibilities for personalized education through the use of interactive recommendation systems in the education domain. However, traditional RL algorithms require extensive interaction with the environment during the training phase. Using unverified recommendation logic in interactions with actual students can give rise to unmanageable problems and hinder effective performance in an educational setting. This is because extrapolation introduces substantial evaluation errors that result in recommendations deviating significantly from the actual educational requirements. To address this limitation, we propose a novel method of offline reinforcement learning called Doubly Constrained deep Q-learning Network (DCQN). This method utilizes two generative models to fit existing student historical interaction data, which in turn, constrains the original policy network to generate new actions based on past interactions, avoiding the occurrence of overestimated actions and reducing extrapolation errors. Empirical results on demonstrate that this approach performs better than existing techniques across D4RL, i.e., datasets for deep data-driven reinforcement learning and real educational datasets.

Leveraging Long Short-Term User Preference in Conversational Recommendation via Multi-agent Reinforcement Learning

Conversational recommender systems (CRS) endow traditional recommender systems with the capability of dynamically obtaining users' short-term preferences for items and attributes through interactive dialogues. There are three core challenges for CRS, including the intelligent decisions for what attributes to ask, which items to recommend, and when to ask or recommend, at each conversation turn. Previous methods mainly leverage reinforcement learning (RL) to learn conversational recommendation policies for solving one or two of these three decision-making problems in CRS with separated conversation and recommendation components. These approaches restrict the scalability and generality of CRS and fall short of preserving a stable training procedure. In the light of these challenges, we tackle these three decision-making problems in CRS as a unified policy learning task. In order to leverage different features that are important to each sub-problem and facilitate better unified policy learning in CRS, we propose two novel multi-agent RL-based frameworks, namely Independent and Hierarchical Multi-Agent UNIfied COnversational RecommeNders (IMA-UNICORN and HMA-UNICORN), respectively. In specific, two low-level agents enrich the state representations for attribute prediction and item recommendation, by combining the long-term user preference information from the historical interaction data and the short-term user preference information from the conversation history. A high-level meta agent is responsible for coordinating the low-level agents to adaptively make the final decision. Experimental results on four benchmark CRS datasets and a real-world E-Commerce application show that the proposed frameworks significantly outperform state-of-the-art methods. Extensive analyses further demonstrate the superior scalability of the MARL frameworks on the multi-round conversational recommendation.

RETRACTED: A Global Structural Hypergraph Convolutional Model for Bundle Recommendation

Bundle recommendations provide personalized suggestions to users by combining related items into bundles, aiming to enhance users’ shopping experiences and boost merchants’ sales revenue. Existing solutions based on graph neural networks (GNN) face several significant challenges: (1) it is demanding to explicitly model multiple complex associations using standard graph neural networks, (2) numerous additional nodes and edges are introduced to approximate higher-order associations, and (3) the user–bundle historical interaction data are highly sparse. In this work, we propose a global structural hypergraph convolutional model for bundle recommendation (SHCBR) to address the above problems. Specifically, we jointly incorporate multiple complex interactions between users, items, and bundles into a relational hypergraph without introducing additional nodes and edges. The hypergraph structure inherently incorporates higher-order associations, thereby alleviating the training burden on neural networks and the dilemma of scarce data effectively. In addition, we design a special matrix propagation rule that captures non-pairwise complex relationships between entities. Using item nodes as links, structural hypergraph convolutional networks learn representations of users and bundles on a relational hypergraph. Experiments conducted on two real-world datasets demonstrate that the SHCBR outperforms the state-of-the-art baselines by 11.07–25.66% on Recall and 16.81–33.53% on NDCG. Experimental results further indicate that the approach based on hypergraphs can offer new insights for addressing bundle recommendation challenges. The codes and datasets have been publicly released on GitHub.

Feature-Alignment-Based Cross-Platform Question Answering Expert Recommendation

Community question answering (CQA), with its flexible user interaction characteristics, is gradually becoming a new knowledge-sharing platform that allows people to acquire knowledge and share experiences. The number of questions is rapidly increasing with the open registration of communities and the massive influx of users, which makes it impossible to match many questions to suitable question answering experts (noted as experts) in a timely manner. Therefore, it is of great importance to perform expert recommendation in CQA. Existing expert recommendation algorithms only use data from a single platform, which is not ideal for new CQA platforms with sparse historical interaction and a small number of questions and users. Considering that many mature CQA platforms (source platforms) have rich historical interaction data and a large amount of questions and experts, this paper will fully mine the information and transfer it to new platforms with sparse data (target platform), which can effectively alleviate the data sparsity problem. However, the feature composition of questions and experts in different platforms is inconsistent, so the data from the source platform cannot be directly transferred for training in the target platform. Therefore, this paper proposes feature-alignment-based cross-platform question answering expert recommendation (FA-CPQAER), which can align expert and question features while transferring data. First, we use the rating predictor composed by the BP network for expert recommendation within the domains, and then the feature matching of questions and experts between two domains by similarity calculation is achieved for the purpose of using the information in the source platform to assist expert recommendation in the target platform. Meanwhile, we train a stacked denoising autoencoder (SDAE) in both domains, which can map user and question features to the same dimension and align the data distributions. Extensive experiments are conducted on two real CQA datasets, Toutiao and Zhihu datasets, and the results show that compared to the other advanced expert recommendation algorithms, this paper’s method achieves better results in the evaluation metrics of MAE, RMSE, Accuracy, and Recall, which fully demonstrates the effectiveness of the method in this paper to solve the data sparsity problem in expert recommendation.

A Hybrid Approach to Trust Node Assessment and Management for VANETs Cooperative Data Communication: Historical Interaction Perspective

Vehicular ad hoc networks (VANETs) provide self-organized wireless multihop transmission, where nodes cooperate with each other to support data communication. However, malicious nodes may intercept or discard data packets, which might interfere with the transmission process and cause privacy leakage. We consider historical interaction data of nodes as an important factor of trust. Thus, this paper focuses on the trust node management of VANETs, which aims to quantify node credibility as an assessment method and avoid assigning malicious nodes. First, the integrated trust of each node is proposed, which consists of the direct trust and the recommended trust. The former is dynamically computed by historical interaction records and Bayesian inference considering penalty factors. The latter defines trust by third-party nodes and their reputation. Second, the process of trust calculation and data communication calls for timeliness. Therefore, we introduce a time sliding window and time decay function to ensure that the latest interaction information has a higher weight. We can sensitively identify malicious nodes and make quick responses. Finally, the experimental results demonstrate that our proposed method outperforms bassline methods, especially with respect to the packet delivery ratio and security.

The Design of Personalized Education Resource Recommendation System under Big Data.

With the advent of the Internet and the era of big data, education is increasingly dependent on data resources to support product and business innovation, and the lack of data resources has severely limited the areas involved. As a general information filtering method, personalized recommendation systems analyze the historical interaction data between users and items to build user interest models in an environment of “information overload”, allowing users to discover and recommend information that interests them. However, the explosive growth of information in the network makes users wander in the sea of information, and it is increasingly difficult to find the information they really need, i.e., information overload. This has given rise to personalized recommendation systems, which currently have more mature applications in industries such as e-commerce, music services, and movie services. To this end, this paper studies and implements a customized educational resource recommendation system that can handle big data. The results show that the values of different similarity calculations all fluctuate with the gradual increase of the number of nearest neighbors, and the algorithm in this paper is maximum at the number of neighbors around 60; then, it is inferred that applying the calculation method to the recommendation algorithm will improve the recommendation accuracy. Therefore, education uses the concept of big data to process the huge amount of education data and find some correlations and laws in education, so as to realize “teaching according to the material, teaching according to the material”.

Enhanced Graph Learning for Recommendation via Causal Inference

The goal of the recommender system is to learn the user’s preferences from the entity (user–item) historical interaction data, so as to predict the user’s ratings on new items or recommend new item sequences to users. There are two major challenges: (1) Datasets are usually sparse. The item side is often accompanied by some auxiliary information, such as attributes or context; it can help to slightly improve its representation. However, the user side is usually presented in the form of ID due to personal privacy. (2) Due to the influences of confounding factors, such as the popularity of items, users’ ratings on items often have bias that cannot be recognized by the traditional recommendation methods. In order to solve these two problems, in this paper, (1) we explore the use of a graph model to fuse the interactions between users and common rating items, that is, incorporating the “neighbor” information into the target user to enrich user representations; (2) the do() operator is used to deduce the causality after removing the influences of confounding factors, rather than the correlation of the data surface fitted by traditional machine learning. We propose the EGCI model, i.e., enhanced graph learning for recommendation via causal inference. The model embeds user relationships and item attributes into the latent semantic space to obtain high-quality user and item representations. In addition, the mixed bias implied in the rating process is calibrated by considering the popularity of items. Experimental results on three real-world datasets show that EGCI is significantly better than the baselines.

Fusion Recommendation System Based on Collaborative Filtering and Knowledge Graph

The recommendation algorithm based on collaborative filtering is currently the most successful recommendation method. It recommends items to the user based on the known historical interaction data of the target user. Furthermore, the combination of the recommended algorithm based on collaborative filtration and other auxiliary knowledge base is an effective way to improve the performance of the recommended system, of which the Co-Factorization Model (<i>CoFM</i>) is one representative research. <i>CoFM</i>, a fusion recommendation model combining the collaborative filtering model <i>FM</i> and the graph embedding model <i>TransE</i>, introduces the information of many entities and their relations in the knowledge graph into the recommendation system as effective auxiliary information. It can effectively improve the accuracy of recommendations and alleviate the problem of sparse user historical interaction data. Unfortunately, the graph-embedded model <i>TransE</i> used in the <i>CoFM</i> model cannot solve the 1-N, N-1, and N-N problems well. To tackle this problem, a novel fusion recommendation model Joint Factorization Machines and <i>TransH</i> Model (<i>JFMH</i>) is proposed, which improves <i>CoFM</i> by replacing the <i>TransE</i> model with <i>TransH</i> model. A large number of experiments on two widely used benchmark data sets show that compared with <i>CoFM</i>, <i>JFMH</i> has improved performance in terms of item recommendation and knowledge graph completion, and is more competitive than multiple baseline methods.

Fusing User Preferences and Spatiotemporal Information for Sequential Recommendation

At present, the research on sequence recommendation mainly focuses on using the historical interaction data between users and items to mine their relationship, so as to predict the next interaction between users and items, then generate the personalized recommendation. Spatiotemporal information is very important to further improve the accuracy and quality of recommendation, but the existing sequence recommendation models are mainly based on recurrent neural network (RNN), and pay less attention to spatiotemporal information. Most of the recommendation models are still in the early stage of merging spatiotemporal context information, and the processing effect of long sequence data is not ideal. a sequential recommendation model integrating user preferences and spatiotemporal information is proposed. The model captures user item long-term preferences through spatiotemporal GRU algorithm and user item short-term preferences through attention mechanism. Finally, the learned long-term and short-term preference features and user portrait features are combined to predict the next recommendation location. The experimental results on two real data sets Foursquare and Brightkite show that the proposed model performs better than state-of-the-arts in three evaluation indicators HR@K, NDCG@K and MAP@K.

Task-oriented Battlefield Situation Information Hybrid Recommendation Model

In the process of interaction between users and battlefield situation information, combat tasks are the key factors that affect users’ information selection. In order to solve the problems of battlefield situation information recommendation (BSIR) for combat tasks, we propose a task-oriented battlefield situation information hybrid recommendation model (TBSI-HRM) based on tensor factorization and deep learning. In the model, in order to achieve high-precision personalized recommendations, we use Tensor Factorization (TF) to extract correlation relations and features from historical interaction data, and use Deep Neural Network (DNN) to learn hidden feature vectors of users, battlefield situation information and combat tasks from auxiliary information. The results are predicted through logistic regression. To solve the multi-source heterogeneity of battlefield situation information, we design a hybrid learning and presentation model that integrates multiple deep learning models such as Doc2Vec, fully connected network and convolutional neural network (CNN), to integrate the rich and diverse data information in situational awareness system effectively. We perform experiments with the maneuvers dataset to test and evaluate the model through scenario simulation.

Incorporating user experience into critiquing-based recommender systems: a collaborative approach based on compound critiquing

Critiques are employed as user feedback in critiquing-based recommender systems and they play an important role in the learning of user preferences, where recommender systems can gradually refine their understanding of user needs and provide better recommendations to users in subsequent interaction sessions. To reduce the effort of user interaction, the advantage of improving the recommendation efficiency by exploring relevant critiquing sessions in the interaction histories of other users has been recognized in recent studies of experience-based critiquing. In this study, we propose a novel approach for processing the historical interaction data in compound critiquing systems. In particular, we describe a history-aware collaborative compound critiquing method, which combines the strategies of preference-based compound critiquing generation and graph-based relevant session identification. Based on a simulation study using real-life data sets, we demonstrated that the proposed method outperformed other experience-based critiquing methods in terms of the recommendation efficiency. We also conducted a retrospective user evaluation, which confirmed the following observations: (1) incorporating user experience into compound critiquing systems significantly improves the performance compared with traditional unit critiquing systems; and (2) our graph-based session identification approach is superior to other baseline methods in terms of reducing the interaction effort of users.