Dynamic User Preferences Research Articles

A Context Awareness Hierarchical Attention Network for Next POI Recommendation in IoT Environment

The rapid increase in the number of sensors in the Internet of things (IoT) environment has resulted in the continuous generation of massive and rich data in Location-Based Social Networks (LBSN). In LBSN, the next point-of-interest (POI) recommendation has become an important task, which provides the best POI recommendation according to the user’s recent check-in sequences. However, all existing methods for the next POI recommendation only focus on modeling the correlation between POIs based on users’ check-in sequences but ignore the significant fact that the next POI recommendation is a time-subtle recommendation task. In view of the fact that the attention mechanism does not comprehensively consider the influence of the user’s trajectory sequences, time information, social relations and geographic information of Point-of-Interest (POI) in the next POI recommendation field, a Context Geographical-Temporal-Social Awareness Hierarchical Attention Network (CGTS-HAN) model is proposed. The model extracts context information from the user’s trajectory sequences and designs a Geographical-Temporal-Social attention network and a common attention network for learning dynamic user preferences. In particular, a bidirectional LSTM model is used to capture the temporal influence between POIs in a user’s check-in trajectory. Moreover, In the context interaction layer, a feedforward neural network is introduced to capture the interaction between users and context information, which can connect multiple context factors with users. Then an embedded layer is added after the interaction layer, and three types of vectors are established for each POI to represent its sign-in trend so as to solve the heterogeneity problem between context factors. Finally reconstructs the objective function and learns model parameters through a negative sampling algorithm. The experimental results on Foursquare and Yelp real datasets show that the AUC, precision and recall of CGTS-HAN are better than the comparison models, which proves the effectiveness and superiority of CGTS-HAN.

A Ranking Recommendation Algorithm Based on Dynamic User Preference.

In recent years, hybrid recommendation techniques based on feature fusion have gained extensive attention in the field of list ranking. Most of them fuse linear and nonlinear models to simultaneously learn the linear and nonlinear features of entities and jointly fit user-item interactions. These methods are based on implicit feedback, which can reduce the difficulty of data collection and the time of data preprocessing, but will lead to the lack of entity interaction depth information due to the lack of user satisfaction. This is equivalent to artificially reducing the entity interaction features, limiting the overall performance of the model. To address this problem, we propose a two-stage recommendation model named A-DNR, short for Attention-based Deep Neural Ranking. In the first stage, user short-term preferences are modeled through an attention mechanism network. Then the user short-term preferences and user long-term preferences are fused into dynamic user preferences. In the second stage, the high-order and low-order feature interactions are modeled by a matrix factorization (MF) model and a multi-layer perceptron (MLP) model, respectively. Then, the features are fused through a fully connected layer, and the vectors are mapped to scores. Finally, a ranking list is output through the scores. Experiments on three real-world datasets (Movielens100K, Movielens1M and Yahoo Movies) show that our proposed model achieves significant improvements compared to existing methods.

Neural Time-Aware Sequential Recommendation by Jointly Modeling Preference Dynamics and Explicit Feature Couplings.

In recommendation, both stationary and dynamic user preferences on items are embedded in the interactions between users and items (e.g., rating or clicking) within their contexts. Sequential recommender systems (SRSs) need to jointly involve such context-aware user-item interactions in terms of the couplings between the user and item features and sequential user actions on items over time. However, such joint modeling is non-trivial and significantly challenges the existing work on preference modeling, which either only models user-item interactions by latent factorization models but ignores user preference dynamics or only captures sequential user action patterns without involving user/item features and context factors and their coupling and influence on user actions. We propose a neural time-aware recommendation network (TARN) with a temporal context to jointly model 1) stationary user preferences by a feature interaction network and 2) user preference dynamics by a tailored convolutional network. The feature interaction network factorizes the pairwise couplings between non-zero features of users, items, and temporal context by the inner product of their feature embeddings while alleviating data sparsity issues. In the convolutional network, we introduce a convolutional layer with multiple filter widths to capture multi-fold sequential patterns, where an attentive average pooling (AAP) obtains significant and large-span feature combinations. To learn the preference dynamics, a novel temporal action embedding represents user actions by incorporating the embeddings of items and temporal context as the inputs of the convolutional network. The experiments on typical public data sets demonstrate that TARN outperforms state-of-the-art methods and show the necessity and contribution of involving time-aware preference dynamics and explicit user/item feature couplings in modeling and interpreting evolving user preferences.

Semantic-enhanced neural collaborative filtering models in recommender systems

Recommendation systems or recommender systems (RSs) are very popular in entertainment websites. With the combination of neural networks and collaborative filtering, Neural Collaborative Filtering (NCF) recommendation methods have shown their outperformance in making item suggestions. However, the lack of semantic relationships between objects makes the NCF unable to capture the complex user–item interactions. Moreover, traditional NCF is unable to capture the dynamic user preference over time. To address these issues, in this paper, we propose novel semantic-enhanced NCF models which are applied to movie rating prediction and movie recommendation. Therefore, MovieLens and IMDB datasets are taken into account as case studies. The proposed models are the integration of ontology-like modeling and deep learning for recommendation tasks into two parts:(1) building the semantic knowledge base for movies and (2) building the user behavior analytic model that has semantic knowledge inference on the knowledge base combined with the sequential preference learned from user sessions, input into the NCF module for making predictions or recommendations. Several experiments have been conducted to show their better recommendation performance than the traditional NCF model.

A multi-intent based multi-policy relay contrastive learning for sequential recommendation.

Sequential recommendations have become a trending study for their ability to capture dynamic user preference. However, when dealing with sparse data, they still fall short of expectations. The recent contrastive learning (CL) has shown potential in mitigating the issue of data sparsity. Many item representations are destined to be poorly learned due to data sparsity. It is better to pay more attention to learn a set of influential latent intents that have greater impacts on the sequence evolution. In this article, we devise a novel multi-intent self-attention module, which modifies the self-attention mechanism to break down the user behavior sequences to multiple latent intents that identify the different tastes and inclinations of users. In addition to the above change in the model architecture, we also extend in dealing with multiple contrastive tasks. Specifically, some data augmentations in CL can be very different. Together they cannot cooperate well, and may stumbling over each other. To solve this problem, we propose a multi-policy relay training strategy, which divides the training into multiple stages based on the number of data augmentations. In each stage we optimize the relay to the best on the basis of the previous stage. This can combine the advantage of different schemes and make the best use of them. Experiments on four public recommendation datasets demonstrate the superiority of our model.

Session-based recommendation with time-aware neural attention network

In situations where user information and detailed knowledge of user behaviors are challenging to obtain, the session-based recommendation is essential. The session-based recommendation (SBR) relies on the current anonymous session sequence to predict the next action. Recently, sequence modeling methods such as recurrent neural networks have been widely used to model user preferences in session-based recommendations. However, the existing SBR methods usually ignore the click time of interacted items, resulting in inaccurate user preference models and poor recommendations. In this article, we propose a time-aware neural attention network to model dynamic user preferences in SBR tasks. First, a global session graph is constructed based on all sessions, and a graph neural network is used to learn the embedding of items. Then, a gated recurrent unit is used to refine the item embedding and capture the general interests of the user based on the current session. Next, a novel time-aware neural attention network is proposed and used to model the main purpose of the user in the session. Finally, the user’s general interests and main purposes are combined to obtain dynamic user preferences for generating recommendations. Extensive experiments show that our method outperforms state-of-the-art session-based recommendation methods significantly and consistently.

Exploiting Long- and Short-Term Preferences for Deep Context-Aware Recommendations

With the tremendous growth in online information, capturing dynamic users’ preferences based on their historical interactions and providing a few desirable items to users have become an urgent service for all businesses. Recurrent neural networks (RNNs) and item-based collaborative filtering (CF) models are commonly used in industries due to their simplicity and efficiency. However, they fail to different contexts that could differently influence current users’ decision-making. Also, they are not sufficient to capture multiple users’ interests based on features of the interacting items. Besides, they have a limited modeling capability for the evolution of diversity and dynamic user preferences. In this article, we exploit long- and short-term preferences for deep context-aware recommendations (LSCAR) to enhance the next item recommendation’s performance by introducing three novel components as follows: 1) the user–contextual interaction module is proposed to capture and differentiate the interaction between contexts and users; 2) the encoded multi-interest module is introduced to capture various types of user interests; and 3) the integrator fusion gate module is used to effectively fuse the related long-term interests to the current short-term part, and the module returns the final user interest representation. Extensive experiments and results for two public datasets demonstrate that the proposed LSCAR outperforms the state-of-the-art models in almost all metrics and could provide interpretable recommendation results.

A novel Sequence-Aware personalized recommendation system based on multidimensional information

Due to the rapid growth of the information overload issue, recommender systems have become necessary and are implemented in numerous facets of human life, including the tourism industry. Today, technological advancements have significantly altered our travels, and these advancements promise even more interactive and exciting experiences in the future. Nowadays, planning and arranging a customized trip well in advance is ideal, as the process can be challenging and time-consuming. This paper proposes a hybrid approach that leverages multidimensional data to enhance personalized trip recommendations while addressing several shortcomings of existing recommender systems, such as their inability to account for dynamic user preferences and diverse contexts. To this end, the proposed method, considering the data sparsity issue, employs a clustering algorithm to reduce the time complexity of discovering Points of Interest (POIs). This approach utilizes user demographic information to address the cold start issue while improving the collaborative filtering (CF) paradigm through an asymmetric schema. Furthermore, this study exploits the context vector model and the Term-Frequency-Inverse-Document-Frequency (TF-IDF) algorithm to present a novel method for quantifying context similarity. Moreover, the method retrieves and ranks a list of candidate routes optimized by applying personalized POIs to sequential travel patterns. Finally, the experimental results demonstrate the superiority of the approach in terms of Precision, Recall, RMSE, MAP, F-Score, and nDCG compared to previous works based on Flickr and STS datasets.

Deep Generative Networks Coupled With Evidential Reasoning for Dynamic User Preferences Using Short Texts

Seeking an efficient solution for the problem of dynamic user preferences on social networks is challenging because the input data are <i>short texts</i> and user preferences usually <i>change</i> over time. This work proposes a novel framework that tackles these challenges based on deep neural networks and the Dempster-Shafer theory of evidence. The framework consists of three primary phases: (1) learning the hidden space of user texts; (2) word generation and mass inference; and (3) mass combination and keyword extraction. In the first phase, user texts are grouped into small batches according to timestamps. Each batch is used for separately training two types of neural networks, the Variational Autoencoder (VAE) and the Generative Adversarial Network (GAN). In the second phase, the generators in the trained VAE and GAN work independently as two <i>experts</i> to generate bunches of tokens for modeling user preferences. Each bunch is considered as one piece of evidence, and is transformed into the so-called mass function in Dempster-Shafer theory by maximum a posterior estimation. In the final phase, Dempster&#x2019;s rule of combination is utilized for fusing the two independent pieces of evidence into an overall mass. This mass is used for extracting top keywords to form the user preferences within a specific time span. The experiments on short text datasets verified that the proposed method outperforms baseline models on many evaluation metrics. Additionally, the output of the proposed framework could be used for visualization, which is useful in many practical applications.

Attention-based dynamic user modeling and Deep Collaborative filtering recommendation

Deep learning (DL) techniques have been widely used in recommender systems for user modeling and matching function learning based on historical interaction matrix. However, existing DL-based recommendation methods usually perform static user preference modeling by using historical interacted items of the user. In this article, we present a time-aware deep CF framework which contains two stages: dynamic user preference modeling based on attention mechanism and matching score prediction based on DL. In the first stage, short-term user preferences are modeled by the time-aware attention mechanism that fully considered the predicted item, the recent interacted items and their interaction time. The resulting short-term preferences are combined with long-term preferences for dynamic user preference modeling. In the second stage, high-order user-item feature interactions are learned by two types of DL models, Deep Matrix Factorization (DMF) and Multiple-Layer Perception (MLP), and the feature interaction vectors of the two models are fused in the last layer of the model to predict the matching score. Extensive experiments on five datasets indicate that our method is superior to the existing time-aware and DL-based recommendation methods in top-k recommendations significantly and consistently.

A novel temporal recommender system based on multiple transitions in user preference drift and topic review evolution

Recommender systems are challenging research problems being exploited to suggest new items or services, such as books, music and movies, and even people, to users based on information about the user profile or the recommended items. To date, collaborative filtering (CF) has become one of the most widely used approaches for recommendations. However, traditional CF methods usually cannot track temporal dynamic user preferences and topic changes to make appropriate suggestions. Moreover, the performance of CF is limited in the case of sparse data. In this paper, we propose a novel temporal recommender system based on multiple transitions in user preference drift, called MTUPD, which employs a multitransition factor and a forgetting time function to investigate the evolution of user preferences. In addition, we consider addressing the rating sparsity issue by using text reviews. Understanding the reviews can facilitate the system grasping whether or not a user is attracted by the appearance of an item and whether the facet of an item’s appearance contributes the most to its ratings. To achieve this, we apply a topic model that automatically classifies hidden topic factors in each time period and incorporate the transition method for both user preferences and relevant review topics. Experiments show that our proposed model outperforms the compared models on eight promising datasets for temporal recommender systems.

A Survey on Session-based Recommender Systems

Recommender systems (RSs) have been playing an increasingly important role for informed consumption, services, and decision-making in the overloaded information era and digitized economy. In recent years, session-based recommender systems (SBRSs) have emerged as a new paradigm of RSs. Different from other RSs such as content-based RSs and collaborative filtering-based RSs that usually model long-term yet static user preferences, SBRSs aim to capture short-term but dynamic user preferences to provide more timely and accurate recommendations sensitive to the evolution of their session contexts. Although SBRSs have been intensively studied, neither unified problem statements for SBRSs nor in-depth elaboration of SBRS characteristics and challenges are available. It is also unclear to what extent SBRS challenges have been addressed and what the overall research landscape of SBRSs is. This comprehensive review of SBRSs addresses the above aspects by exploring in depth the SBRS entities (e.g., sessions), behaviours (e.g., users’ clicks on items), and their properties (e.g., session length). We propose a general problem statement of SBRSs, summarize the diversified data characteristics and challenges of SBRSs, and define a taxonomy to categorize the representative SBRS research. Finally, we discuss new research opportunities in this exciting and vibrant area.

Attention-based dynamic user preference modeling and nonlinear feature interaction learning for collaborative filtering recommendation

The traditional collaborative filtering (CF) method based on static user preference modeling and linear matching function learning severely limits the recommendation performance. To solve the above problem, in this article, we adopt dynamic user preference modeling and nonlinear matching function learning in the CF recommendation. For dynamic user preference modeling, a two-layer neural attention network is used, which fully considers the predicted item, the recent historical interacted items and their interaction time to estimate the contribution weight of each interacted item in user preferences modeling. For nonlinear matching function learning, we add a single hidden layer neural network on top of the traditional matrix factorization (MF) model, which can significantly improve the feature interaction learning capabilities of the model with only a few additional parameters. Extensive experiments show that our method significantly outperforms the state-of-the-art CF methods and the key technologies we proposed in this research have a positive effect on improving the recommendation performance.

ADCF: Attentive representation learning and deep collaborative filtering model

In this paper, we propose a deep collaborative filtering recommendation model, which consists of an attention-based representation learning component and a multi-input matching function learning component. This model takes interaction matrix based on implicit feedback as data source to construct representations of long-term user preferences and item latent features. In the representation learning, a time-aware attention network is used, which uses the embedding vectors of the predicted item, recent historical interaction items, and the interaction time of recent historical interaction items to estimate the weights of different historical interaction items to short-term user preferences modeling. Then, the dynamic user preference representation can be obtained by combining short-term preferences with long-term preferences. In the matching function learning, a multi-input deep learning model is used. Its input includes not only the dynamic user preference representation and the item latent feature representation, but also the linear interaction between the two representations, so that the model has more powerful feature interactions learning ability. Experimental results on four datasets from different domains show that our method is largely superior to the state-of-the-art collaborative filtering methods, and the novel techniques we propose in this paper are effective in improving recommendation performance.

Searching for Flexibility in Corporate Real Estate Portfolio: Six Co-Working Strategies for User Corporations

The increasing competitive pressures and dynamic user preferences have resulted in a fast-paced and uncertain business environment. In the face of these circumstances, organizations are looking into alternatives to incorporate flexibility to become more adaptive and responsive to change. In this line, co-working, typically associated with freelancers, entrepreneurs, and startups, has become a particularly interesting alternative in the market that has caught the attention of corporate occupiers. Therefore, the aim of this paper is to identify co-working strategies that can be implemented as part of the corporate real estate portfolio, in alignment with the flexibility demands of the organization. This nascent research topic is studied through 5 qualitative case studies including in-depth, semi-structured interviews with corporate real estate managers and related case documentation. The results evidence the different motivations that the organizations have when incorporating co-working in their property portfolio. As seen across the cases, organizations in different stages of maturity are implementing co-working as the main office location or as a temporary or complementary space solution, through six different strategies: (1) Swing Space, (2) Expansion Space, (3) Core and Flex, (4) Touchdown Space, (5) Testing Market, and (6) Temporary Projects and Staff. This research evidences that each strategy plays a specific role in the corporate real estate portfolio and implies different sources of flexibility that support the physical, functional, and financial flexibility demands of the organization.

HTDA: Hierarchical time-based directional attention network for sequential user behavior modeling

Personalized recommendation is one of the most effective ways to alleviate the problem of information explosion. Recurrent Neural Network (RNN) sequential recommendation based on user preference modeling has been widely investigated recently. In this paper, a Hierarchical Time-based Directional Attention (HTDA) network is proposed to enhance sequential recommendation by applying fine-grained user intention representation and dynamic user preference representation with rich global sequential interaction features. Our analyses and results significantly differ from earlier related works in two aspects: 1) We put forward a session-level representation method, based on multi-dimensional attention mechanism, to enhance the degree of matching between the user interaction sequence and the user intention and reduce the impact of noise interaction; 2) We propose to integrate a time-based directional attention mechanism into RNN, to capture the sequential patterns of interaction sessions more effectively and improve the understanding of user dynamic preferences. The main contribution of our work is to effectively combine the accurate description of user intentions and the powerful expression of user dynamic preferences through a hierarchical attention architecture. In turn, our architecture can provide a more favorable framewrok for further enhancing the recommendation results. Extensive experiments are conducted on two public benchmark datasets, and very promising performance can be achieved with our proposed HTDA-based framework.

A Smartphone-Based Technique to Detect Dynamic User Preferences for Tailoring Behavioral Interventions: Observational Utility Study of Ecological Daily Needs Assessment.

BackgroundMobile health apps are promising vehicles for delivering scalable health behavior change interventions to populations that are otherwise difficult to reach and engage, such as young adults with psychiatric conditions. To improve uptake and sustain consumer engagement, mobile health interventions need to be responsive to individuals’ needs and preferences, which may change over time. We previously created an ecological daily needs assessment to capture microprocesses influencing user needs and preferences for mobile health treatment adaptation.ObjectiveThe objective of our study was to test the utility of a needs assessment anchored within a mobile app to capture individualized, contextually relevant user needs and preferences within the framework of a weight management mobile health app.MethodsParticipants with an iOS device could download the study app via the study website or links from social media. In this fully remote study, we screened, obtained informed consent from, and enrolled participants through the mobile app. The mobile health framework included daily health goal setting and self-monitoring, with up to 6 daily prompts to determine in-the-moment needs and preferences for mobile health–assisted health behavior change.ResultsA total of 24 participants downloaded the app and provided e-consent (22 female; 2 male), with 23 participants responding to at least one prompt over 2 weeks. The mean length of engagement was 5.6 (SD 4.7) days, with a mean of 2.8 (1.1) responses per day. We observed individually dynamic needs and preferences, illustrating daily variability within and between individuals. Qualitative feedback indicated preferences for self-adapting features, simplified self-monitoring, and the ability to personalize app-generated message timing and content.ConclusionsThe technique provided an individually dynamic and contextually relevant alternative and complement to traditional needs assessment for assessing individually dynamic user needs and preferences during treatment development or adaptation. The results of this utility study suggest the importance of personalization and learning algorithms for sustaining app engagement in young adults with psychiatric conditions. Further study in broader user populations is needed.

Time-aware user profiling from personal service ecosystem

The tremendous growth of services available on the Internet makes user profiling an increasing important topic in web personalization. While profiling a user, one of the main challenges is that user preferences and interests usually change over time. It would be much appreciated, if the user profile could describe these changes. Such user profile is defined as time-aware user profile (TUP) in this paper. The TUP can delineate both dynamic user preferences and interests and evolutions of them, which illustrates that it can represent users more clearly and accurately. A TUP consists of a set of time slot-specific user profiles, which is defined as sub-time-aware user profile (sTUP). And a TUP is special for a user; then, its sTUPs separately represent preferences and interests of this user for different time slots. Constructing a TUP needs the constructions of its sTUPs. While constructing a sTUP from the user behavior logs for a time slot, in this paper, we present a novel learning model called personal service ecosystem (PSE) to delineate user preferences and interests naturally. Different from PSE, sTUP is the representation model. For a time slot, the sTUP is constructed based on the PSE recovered from the user behavior logs of this time slot. In terms of a TUP, two main approaches to analyze the evolutions of sTUPs in this TUP are further discussed. They are the evolution analysis of various margins between every two neighboring sTUPs called EA-VM and the evolution analysis of different sTUPs on raw data defined as EA-RD. EA-VM mainly focuses on the fine-grained evolution between two neighboring time slots by analyzing these various margins. It can be further divided into the clustering method and the distribution analysis approach. EA-RD chiefly pays attentions to the overall evolution of such sTUPs. By a comprehensive survey on the raw data of sTUPs, we summarize six interesting overall evolution patterns, including refugee pattern, periodic pattern, stable pattern, fluctuant pattern, emergency pattern and zombie pattern. Eventually, a systematic empirical study based on our collection data has been conducted, where smartphone users are taken as examples to illustrate our proposed models and approaches. And experimental results highlight the superiority of these approaches.

A graph embedding based model for fine-grained POI recommendation

Point-of-interest (POI) recommendation is an important technique widely used in self-driving services. While POI recommendation aims to recommend unvisited POIs to self-driving users, users always expect their intended items can be suggested together with these POIs, e.g. what activities to perform at the recommended places. However, existing methods cannot well support such POI recommendation in a finer granularity. In this paper, we investigate this new problem and propose a novel POI-based item recommendation model via graph embedding. The model accurately captures the joint effect of geographical and temporal influences on both POI-level and item-level recommendation in a shared space, which can address data sparsity and cold start problems effectively. To optimize the model efficiently and accurately, a novel weighted negative sampling strategy is designed. Besides, we propose a novel fine-grained user dynamic preference modeling method, which can accurately capture dynamic user preferences in a finer granularity based on the embeddings of both POIs and items. Comprehensive experimental studies have been conducted on three datasets. Results show that our model achieves significant improvement over state-of-the-art baselines.

Restricted Boltzmann Machine-driven Interactive Estimation of Distribution Algorithm for personalized search

Effective and efficient personalized search is one of the most pursued objectives in the era of big data. The challenge of this problem lies in its complex quantifying evaluations and dynamic user preferences. A user-involved interactive evolutionary algorithm is a good choice if it has reliable preference surrogate and powerful evolutionary strategies. A Restricted Boltzmann Machine (RBM) assisted Interactive Estimation of Distribution Algorithm (IEDA) is presented to enhance the IEDA in solving the personalized search. Specifically, a dual-RBM module is developed to simultaneously provide a preference surrogate and a probability model for conducting the individual selection and generation of the IEDA. Firstly, the positive and negative preferences of the currently involved user in IEDA are distinguished and combined to achieve a dual-RBM, and then the weighted energy functions of the RBM model together with social group information from users with similar preferences are designed as the preference surrogate. The probability of the trained positive RBM on the visible units is fetched as the reproduction model of EDA since it reflects the attribute distributions of more preferred items. Some benchmarks from the Movielens and Amazon datasets are applied to experimentally demonstrate the superiority of the proposed algorithm in improving the efficiency and effectiveness of the interactive evolutionary computations served personalized search.