Addressing itinerary modification is crucial for enhancing the travel experience as it is a frequent requirement during traveling. However, existing research mainly focuses on fixed itinerary planning, leaving modification underexplored due to the scarcity of shape need-to-modify itinerary data. To bridge this gap, we formally define the itinerary modification task and propose a general pipeline to construct the corresponding dataset, namely iTIMO. This pipeline frames the generation of shape need-to-modify itinerary data as an intent-driven perturbation task. It instructs large language models to perturb real-world itineraries using three operations: REPLACE, ADD, and DELETE. Each perturbation is grounded in three intents: disruptions of popularity, spatial distance, and category diversity. Furthermore, hybrid metrics are proposed to ensure perturbation effectiveness. We conduct comprehensive benchmarking on iTIMO to analyze the capabilities and limitations of state-of-the-art LLMs. Overall, iTIMO provides a comprehensive testbed for the modification task, and empowers the evolution of traditional travel recommender systems into adaptive frameworks capable of handling dynamic travel needs. Dataset, code and supplementary materials are available at https://github.com/zelo2/iTIMO.

ABSTRACT Rapidly growing social media has become a key platform in recent years for influencing public sentiment and shaping hot public opinion. Against this backdrop, accurately predicting social users' opinions holds significant importance for public opinion analysis and guidance. Existing research generally focuses on the macro level of hot events. However, it fails to fully exploit the interaction data between high‐impact users and ordinary users. Moreover, such studies often neglect individual user characteristics and behavioural differences, which limits their capability in accurately forecasting user opinion trends. This paper proposes a novel approach to predict social users' opinions by leveraging high‐impact users' interaction data. The approach utilises the posts and interactive comments of high‐impact users to construct multiple user–post opinion matrices and employs block‐wise matrix factorisation to extract the latent embeddings of users and posts. Building upon this, the historical comment–post pairs are encoded using BERT, and a bidirectional cross‐attention mechanism is applied to model the semantic correlations between comments and posts, yielding cross‐attentional pair representations. To integrate these representations, a dynamic attention mechanism is used to weight and aggregate historical behaviours, generating an attention‐weighted semantic representation that is highly relevant to the current task. Finally, the multi‐source features of users and posts are fused through a transformer architecture and fed into a multi‐layer perceptron for opinion prediction. Experiments on a real‐world dataset demonstrate that the proposed approach achieves superior performance in user opinion prediction, showing promising potential for practical applications.

It is challenging to achieve an optimal worker allocation for production lines of large-scale industrial enterprises due to the complex requirements for workers’ capabilities. Some optimization approaches have been proposed to solve this problem from different perspectives. However, they have not fully explored the multidimensional features of both workers and production lines, making it hard to obtain an optimal allocation. This article proposes a novel approach to collaborative allocation optimization of production line workers by incorporating multidimensional feature measurement and the environment-classes, agents, roles, and objects (E-CARGO) model. First, we develop a comprehensive evaluation system to quantify the diverse features of workers and production lines. Based on it, an adaptability assessment mechanism is designed to measure the matching degree of workers for different production lines. Afterward, the role-based collaboration theory and the E-CARGO model are innovatively utilized to formalize the worker allocation problem. Meanwhile, the key constraints are identified to guarantee the reasonability of allocation, and an efficient solution via CPLEX package is proposed. Finally, the case analysis and simulation experiments verify the effectiveness of the proposed approach.

Timely and effective early warning is essential for proactive intervention to mitigate students’ learning risks. Existing studies on learning early warning primarily predict students’ knowledge mastery based on academic performance. However, they lack the assessment of course development objective achievement. According to the outcome-based education (OBE) concept, the course objective achievement serves as the foundation for comprehensive student assessment spanning knowledge acquisition, learning ability, and learning attitude. Using the course objective achievement as a guide for learning early warning can help to obtain more objective and accurate warning results. This article proposes a novel approach to learning early warning guided by course objective achievement via knowledge state and learning state modeling. This approach constructs knowledge states related to course objectives through a deep knowledge tracing model and derives the learning states comprising learning ability and learning attitude from multidimensional learning behavior data. The achievement state, fusing the knowledge and learning states, is then fed into a transformer model to capture the temporal dynamics of the achievement state and predict the achievement levels for each course objective. Based on these predictions, a four-level warning rule is employed to assess students’ learning risks. Experiments based on two real-world datasets demonstrate the effectiveness and superiority of the proposed approach. The approach provides a new research paradigm and a feasible solution to achieve accurate personalized early learning warning.

In recent years, group self-driving tours have become a popular tour mode among Chinese tourists. However, current research on tourism resource recommendation generally overlooks the distinctive tourism styles and group preferences of self-driving tourists. To address the diversity in tourism styles and group preferences among self-driving tourists, this paper proposes a tourism resource recommendation method tailored for self-driving tours, based on tourism style modeling and group decision-making. First, by integrating tourist and tourism resource information, a knowledge graph of the self-driving tour environment is constructed, enhancing the accuracy of identifying similar tourists. Next, seven fundamental tourism types are incorporated along with factors specific to self-driving tours to better model the tourism styles of tourists, thereby improving recommendation accuracy. Furthermore, to accommodate the varied preferences within a self-driving group, a non-compensatory group decision-making strategy is employed to ensure the recommendation results meet the needs of all group members. Finally, case studies validate the effectiveness of the proposed method. This study can provide a new perspective for resource recommendation in self-driving tours.

With the acceleration of globalization, the targeted international communication of national images contributes to enhancing a nation’s soft power and international recognition. It is challenging to select appropriate resources from the mass candidates for creating promotional works of national image. Existing research only focuses on the methodologies and lacks the systematic modeling and solving of national image resources recommendation. A collaborative recommendation approach to national image resources is proposed for targeted international communication. In it, the multidimensional features of national image resources and characteristics of communication audiences are modeled, and an evaluation mechanism is proposed to measure the comprehensive compatibility between national image resources and communication audiences. By innovatively introducing the role-based collaboration (RBC) theory and the environment-classes, agents, roles, groups, and objects (E-CARGO) model, the national image resources recommendation is formalized as a collaborative optimization problem. The mathematical model is built and solved via an optimization package. Finally, the case study and experiments show that the approach is efficient, feasible, and conducive to enhancing the efficiency of selecting national image resources. It offers a novel research paradigm for targeted international communication.

Teaching early warning is of great significance for avoiding teaching risks and continuously improving teaching quality. However, none of existing approaches assess the degree of course goals attainment and teacher's teaching quality from the perspective of cognitive diagnosis. This poses a challenge in providing accurate teaching early warning. This paper proposed an early warning approach for teachers based on cognitive diagnosis and long short-term memory (LSTM). First, this approach accurately evaluates students' cognitive status on knowledge concepts using a cognitive diagnosis model to assess their knowledge understanding degree and knowledge application ability. Second, the cognitive status on knowledge concepts is utilized to assess students' attainment degree of course goals and teachers' teaching quality. Third, the teachers' teaching quality is predicted in the future by using the LSTM network to mine students' learning process data, Finally, an accurate teaching early warning is provided to teachers based on a four-level early warning evaluation rule. In experiments, the real datasets are used and the results reveal that the proposed approach can accurately diagnose students' cognitive status and effectively predict teachers' teaching quality. This approach can provide an accurate teaching early warning service for teachers.

Recently, the city road trips have become one of the mainstream travel styles for Chinese tourists. Existing research does not comprehensively consider tourists' subjective preferences for sightseeing, dining, and accommodation, and analyze the objective constraints related to city road trips. It is difficult for tourists to obtain route planning solutions with high satisfaction of city road trips. A route planning approach for city road trips is proposed to meet these preferences and constraints. This approach defines two types of POIs (points of interest), i.e., attractions and rest spots, and identifies key constraints possibly affecting POI selections. Based on them, the tourist’s route for one day is divided into three sub-routes. Each sub-route consists of two different POIs including an attraction and a rest spot. First, an appropriate attraction is selected as the center of a sub-route. Second, a suitable rest spot is assigned to each sub-route according to its center. The E-CARGO model is utilized to formalize the route planning problem of city road trips and an effective solution is provided. Case study and simulation experiments show that the approach is efficient and feasible to achieve the maximum tourist satisfaction in city road trips.

It is important to select appropriate national image resources to build a nation’s images for targeted international communication of national images. Existing research only focusing on the methodologies, lacks the systematic modeling and solving of national image resources recommendation. A collaborative recommendation approach to national image resources is proposed. In the approach, an evaluation model of national image resources and an evaluation model of communication audiences are put forward, and an evaluation mechanism is proposed to measure the comprehensive compatibility between national image resources and communication audiences. By innovatively introducing the role-based collaboration (RBC) theory and the environment-classes, agents, roles, groups, and objects (E-CARGO) model, the national image resources recommendation is formalized as a collaborative optimization problem. The mathematical model is built and solved via an optimization package. Finally, the case study and experiments show that the approach is efficient, feasible, and conducive to enhancing the efficiency of national image resources recommendation. It offers a novel research paradigm for targeted international communication of national images.

With the massive popularization of e-learning, collaborative learning via learning teams has become indispensable to enhancing the learning efficiency and learning quality of overall learners. The team leader usually plays a key role in collaborative learning. However, the existing research ignores the key characteristics of learners and constraints relevant to e-learners when identifying appropriate team leaders and compatible members. A novel collaborative optimization approach to learning team formation is proposed based on a refined learner model and the environments—classes, agents, roles, groups, and objects (E-CARGO) model. With the proposed approach, a learner is modeled by combining 5-D characteristics (i.e., cognitive ability, leadership, sociability, learning style, and personality) and three types of constraints (e.g., conflicts, genders, and the number of members), and an assessment mechanism is designed to measure the comprehensive abilities of learners for identifying an ideal team leader and selecting the team members for a team. By innovatively introducing the role-based collaboration theory and E-CARGO model, the leader-centered learning team formation problem is formalized as a collaborative optimization problem. The mathematical model and the constraint relations are established for this problem, which is solved based on the IBM CPLEX package. Finally, a case study and experiments demonstrate that the proposed approach is efficient and feasible, in favor of improving the satisfaction degree of learners.

Recently, the road trip in regional central cities has become one of the mainstream travel styles for Chinese tourists. However, existing research fails to plan the road trip route in Chinese regional central cities due to its flexibility, complexity, and long-term characteristic. A collaborative route planning approach for road trips, namely CoRPoRT, is proposed. This approach decomposes the route planning of a road trip into two role-based collaboration subproblems to alleviate the complexity issue. Then, the environments–classes, agents, roles, groups, and objects (E-CARGO) model is innovatively employed to formalize the problems and guide the optimization process for dealing with the long-term characteristic of road trips. Moreover, we identify the complex constraints affecting point of interest selections and propose an efficient solution via the IBM CPLEX solver to handle the flexibility of road trips. Finally, a case study and simulation experiments verify the effectiveness of CoRPoRT. CoRPoRT presents a general problem modeling method and a novel research paradigm for the route planning of road trips.

Learning early warning is of great significance for coping with students' learning risks. The existing research fails in modeling the fluctuation of students' learning states and providing the multi-level early warning for students at different levels. To address them, a new approach of learning early warning is proposed to predict at-risk students in e-learning environment by combining cognitive diagnosis with learning behaviors analysis. In this approach, the students' learning process is modeled from four dimensions, i.e., learning quality, learning engagement, latent learning state, and historical learning performance. The convolutional neural network and long short-term memory network are used to explore the students' latent learning features. Then, the Adaboost algorithm is applied to predict students' learning performance. Based on the predicted performance, the evaluation rules are designed to provide multi-level learning early warning for students. Finally, the experiments demonstrate that the proposed method could predict at-risk students efficiently and accurately.

The prediction of examinee performance via cognitive diagnosis models might provide an important decision-making support for personalized learning instruction in an e-learning system. Aiming at the uncertainty of learners' skill proficiency caused by the complexity of skills, and the large-scale volume of score profiles, a collaborative prediction approach of examinee performance is proposed based on a new fuzzy cloud cognitive diagnosis model. In this approach, the normal cloud models are used to measure the uncertainty of the skill proficiency from three aspects (i.e., expectation, variation degree, and variation frequency), and an e-learner’s skill proficiency is characterized with a fuzzy interval number. Based on a collaborative parameter estimation method, the predicted scores on every test item could be obtained for learners. Finally, the experiments demonstrate that this approach provides good accuracy and less execution time for predicting examinee performance than other approaches.

Facing with the massive learning resources, the learners are often confronted with information disorientation and overload problems. To help learners select the appropriate MOOCs quickly, a hybrid recommendation approach of personalized MOOC resources is proposed by exploiting the user context to capture the learners' explicit and implicit features. An improved hybrid similarity calculation method is presented to identify the neighboring users for reducing the calculation errors, and an improved course modeling method is used to extract semantic information for enhancing the accuracy of course modeling with low labor cost. Based on this approach, a real system is developed. The experiments demonstrate that this approach provides the higher recommendation accuracy and ideal execution performance compared with the traditional approaches for supporting personalized learning efficiently.

It is a fundamental function of a personalized elearning system to recommend suitable exercises to learners for improving their learning efficiencies and qualities. These exercises relevant to the current learning progress and the skill proficiency of learners could be selected by analyzing their score profiles in the past exams. Aiming at the limitations of existing research, a new exercise recommendation approach is proposed based on cognitive diagnosis and Neutrosophic set. In it, the learners' cognitive status is measured from multiple perspectives comprehensively by introducing the Neutrosophic set theory. The similarity between the learners is calculated with a Neutrosophic set method. The learner's performance on the new exercises could be predicted by collaborative filtering algorithm, and the exercises suitable to learners are recommended to them according to their preferences. The experiments show that the accuracy of proposed approach is higher than the existing approaches.

As the number of cloud services (CSs) offering similar functionality is growing, more attention has been payed on the quality of service (QoS) of CSs. However, in a dynamic cloud environment, the explicit and inherent variation of QoS causes the single CS selection via collaborative filtering techniques (CSS-CFT) to be challenging. A variation-aware approach via collaborative QoS prediction is proposed to select an optimal CS according to users’ non-functional requirements. Based on time series QoS data, this approach utilizes a set of specific cloud models to quantify the variation characteristics of QoS from the four aspects including central tendency, variation range, frequency of variation and period. To exactly identify the neighboring users for a current user, this paper employs the double Mahalanobis distances to measure the similarity of QoS cloud models. The variation-aware CSS-CFT is formulated as a multi-criteria decision-making problem, and an improved TOPSIS method is exploited to solve it, by considering both the objective QoS variation and subjective user preferences during different time periods. The experiments based on a real-world dataset demonstrate that the proposed approach can enhance the accuracy of CSS-CFT in a high-variance environment without noticeable increase of selection time, in comparison to the existing approaches.

Recently, growing cloud services (CSs) have been leased by organizations for high-performance computation and massive data storage of data-intensive applications (DiAs). To improve the resource utilization of leased CSs, it has become a challenging task to optimize infrastructure as a service CS composition for DiAs (ICSCDs) from the user side. This paper proposes a resource utilization-aware collaborative optimization approach. Targeting the collaboration features of tasks in a DiA, the environments-classes, agents, roles, groups, and objects model is used to formalize the ICSCD problem from the perspective of role-based collaboration. Aiming at the dynamic characteristics of the cloud environment, an integrated method is presented to evaluate the qualification of CSs via the interval numbers with multiple parameters. Based on the exact qualification values, the ICSCD can be optimized for improving the resource utilization of the CSs. A solution using the IBM ILOG CPLEX optimization package is put forward to solve the problem. The experimental results demonstrate that the approach can provide high precision, performance, stability, resource utilization, and low usage cost for the resource utilization-aware ICSCD from the user side.

The multi-valued evaluations of quality of service (QoS), the complicated constraints between cloud services (CSs) and the collaborative resource assignments add many difficulties to the problem of CS composition for data-intensive applications (DiA) in a hybrid cloud (CSCD-HC). Solving the CSCD-HC problem has become a challenging task due to the uncertain QoS, the diverse hardware configurations and the flexible pricing about CSs. This paper proposes a collaborative optimization approach for CSCD-HC. This approach models a DiA as a role-based collaboration (RBC) system and employs the environments-classes, agents, roles, groups, and objects (E-CARGO) model to formalize the CSCD-HC problem with complicated constraints. To deal with the multi-valued QoS evaluations, this paper exploits the cloud model theory to analyze the performance of CSs, and presents a new method utilizing the Mahalanobis distance to improve the similarity calculation of QoS cloud models. Based on it, the qualification of candidate CSs can be precisely measured for supporting CS composition. A solution via the IBM ILOG CPLEX optimization package is put forward to solve the CSCD-HC problem. The experimental results demonstrate that the proposed approach is effective and feasible for optimizing CSCD-HC.

With the growing cloud services (CSs) rented by an organization, it has become a challenging problem to optimize the CS composition for data-intensive applications (DiAs) from the user side, with consideration of improving the resource utilization of rented CSs. This paper proposes a resource utilization-aware approach to optimizing the CS composition for DiAs (CSCD). From the perspective of role-based collaboration, this approach utilizes the environments - classes, agents, roles, groups, and objects (E-CARGO) model to formalize the CSCD problem. The qualification of a CS for one task is assessed and the compatibility between new tasks and the running task is identified. A solution using IBM ILOG CPLEX package is put forward to optimize the CSCD problem. The experimental results demonstrate that the proposed approach is effective and feasible for optimizing the resource utilization-aware CSCD problem from the user side.

In a team of collectivism, the team performance or the team interest is emphasized. Many individuals may find that the team interest is not always consistent with their own. Assigning positions to team members is one such scenario. An individual may be assigned to a position that is not preferred. Is there any policy for an individual to apply in order to get the preferred position? Role-Based Collaboration (RBC) is a methodology that advocates collectivism. The Environments - Classes, Agents, Roles, Groups, and Objects (E-CARGO) is a good tool to investigate this issue. The contributions of this work include: 1) a data analysis method for a person to pursue the preferred position (role) in a team based on the E-CARGO model and Group Role Assignment (GRA) algorithm; 2) a set of experiments for the proposed methods; and 3) the confirmation of a list of common sense principles with the support of experiments.

Owing to the deficiency of usage experiences and the information overload of QoE (quality of experience) evaluations from consumers, how to discover the trustworthy cloud services is a challenge for potential users. This paper proposed a cloud service recommendation approach based on trust measurement using ternary interval numbers for potential user. The concept of ternary interval number is introduced. The user feature maybe affecting the QoE evaluations are analyzed and the client-side feature similarity between consumers and potential user is calculated. The transform mechanism from trust evaluations to ternary interval number is presented by employing the K-means clustering algorithm. On the basis of multi-attributes trust aggregation based On FAHP (fuzzy analytic hierarchy process) method, a new possibility degree formula is designed for ranking ternary interval numbers and selecting trustworthy service. Finally, the experiments and results show that this approach is effective to improve the accuracy of the trustworthy service recommendation.