Patent•
Learning user preferences using sequential user behavior data to predict user behavior and provide recommendations
10 Nov 2016-
TL;DR: In this paper, a machine learning algorithm is used to predict a subsequent action of a user among the various users based on the various clusters of prior actions taken by multiple users, and then a new action is recommended to a new user based on a cluster associated with the new user.
Abstract: Certain embodiments involve learning user preferences and predicting user behavior based on sequential user behavior data. For example, a system obtains data about a sequence of prior actions taken by multiple users. The system determines a similarity between a prior action taken by the various users and groups the various users into groups or clusters based at least in part on the similarity. The system trains a machine-learning algorithm such that the machine-learning algorithm can be used to predict a subsequent action of a user among the various users based on the various clusters. The system further obtains data about a current action of a new user and determines which of the clusters to associate with the new user based on the new user's current action. The system determines an action to be recommended to the new user based on the cluster associated with the new user. The action can include a series or sequence of actions to be taken by the new user. The system further provides the series or sequence of actions or an action of the series or sequence to the new user.
Citations
More filters
Patent•
14 Jun 2018TL;DR: In this paper, a system includes at least one electronic processor configured to access data representing historical tasks performed by a user through one user device, determine, based on the data, a first plurality of tasks associated with a life event of the user, and store an association between the first plurality and the life event.
Abstract: Methods and systems for providing digital assistance. One system includes at least one electronic processor configured to access data representing historical tasks performed by a user through at least one user device, determine, based on the data, a first plurality of tasks associated with a life event of the user, and store an association between the first plurality of tasks and the life event. The electronic processor is also configured to, in response to a current occurrence of the life event experienced by the user, retrieve the association and generate a user interface for display to the user, the user interface including a second plurality of tasks for addressing the current occurrence of the life event based on the first plurality of tasks.
1 citations
Patent•
05 Nov 2019
TL;DR: In this article, an intelligent financial management recommendation method based on user continuous behavior sequence characteristics is presented. But the method comprises the steps of obtaining financial product attribute data and user attribute data, performing data cleaning and standardized management, classifying the data and performing transformer model architecture training on the data, inputting information intoa trained model, outputting personal financial recommendation by the model, recommending the personal financial recommender to a user through multiple channels, and obtaining feedback information.
Abstract: The invention discloses an intelligent financing recommendation method based on user continuous behavior sequence characteristics. The method comprises the steps of obtaining financial product attribute data and user attribute data, performing data cleaning and standardized management, classifying the data, performing transformer model architecture training on the data, inputting information intoa trained model, outputting personal financial recommendation by the model, recommending the personal financial recommendation to a user through multiple channels, and obtaining feedback information.In the present invention, according to the intelligent financial management recommendation method, the feature extraction is carried out on user information data uploaded by a user terminal. Particularly, a sequence signal of a user behavior sequence is captured by using a Transformer new architecture model. The previous behavior sequence characteristics of the user are combined, so that the accuracy of recommending financial products to the user is improved. Therefore, the click rate and the purchase rate of the financial product by the user are automatically improved. The intelligent financial recommendation system can be safely used. The knowledge degree between the user and the bank financial product is enhanced. The mutual benefit and mutual benefit effects are achieved for the user and the bank.
Patent•
07 Aug 2020
TL;DR: In this paper, the authors describe methods and apparatuses for automated predictive product recommendations using reinforcement learning, where a server generates a context vector for each user, the context vector comprising a multidimensional array corresponding to historical activity data and assigns each context embedding to an embedding cluster.
Abstract: Methods and apparatuses are described for automated predictive product recommendations using reinforcement learning. A server captures historical activity data associated with a plurality of users. The server generates a context vector for each user, the context vector comprising a multidimensional array corresponding to historical activity data. The server transforms each context vector into a context embedding. The server assigns each context embedding to an embedding cluster. The server determines, for each context embedding, (i) an overall likelihood of successful attempt and (ii) an incremental likelihood of success associated products available for recommendation. The server calculates, for each context embedding, an incremental income value associated with each of the likelihoods of success. The server aggregates (i) the overall likelihood of successful attempt, (ii) the likelihoods of success, and (iii) the incremental income values into a recommendation matrix. The server generates instructions to recommend products based upon the recommendation matrix.
Patent•
22 Mar 2019
TL;DR: In this article, a lottery user activity prediction method is proposed, which comprises of the following steps: acquiring original user data, extracting and converting the original users' data, classifying and loading the original user's data into a database in a specified format, and finally loading the user's original data into the database in the specified format.
Abstract: The invention discloses a lottery user activity prediction method, which comprises the following steps: acquiring original user data; extracting and converting the original user data; classifying andloading the original user data into a database in a specified format; and loading the original user data into a database in a specified format. Preprocessing the original user data stored in the database to obtain multi-dimensional user data; Obtaining a prediction feature set related to user activity according to the multi-dimensional user data; inputting The prediction feature set into a pre-trained GBDT algorithm-based activity prediction model to predict user activity. Correspondingly, the invention also discloses a lottery user activity prediction system, a terminal device and a computer-readable storage medium. The technical proposal of the invention can reduce the prediction difficulty of the lottery user activity and improve the prediction accuracy.
References
More filters
28 Oct 2010
TL;DR: This handbook illustrates how recommender systems can support the user in decision-making, planning and purchasing processes, and works for well known corporations such as Amazon, Google, Microsoft and AT&T.
Abstract: The explosive growth of e-commerce and online environments has made the issue of information search and selection increasingly serious; users are overloaded by options to consider and they may not have the time or knowledge to personally evaluate these options. Recommender systems have proven to be a valuable way for online users to cope with the information overload and have become one of the most powerful and popular tools in electronic commerce. Correspondingly, various techniques for recommendation generation have been proposed. During the last decade, many of them have also been successfully deployed in commercial environments. Recommender Systems Handbook, an edited volume, is a multi-disciplinary effort that involves world-wide experts from diverse fields, such as artificial intelligence, human computer interaction, information technology, data mining, statistics, adaptive user interfaces, decision support systems, marketing, and consumer behavior. Theoreticians and practitioners from these fields continually seek techniques for more efficient, cost-effective and accurate recommender systems. This handbook aims to impose a degree of order on this diversity, by presenting a coherent and unified repository of recommender systems major concepts, theories, methodologies, trends, challenges and applications. Extensive artificial applications, a variety of real-world applications, and detailed case studies are included. Recommender Systems Handbook illustrates how this technology can support the user in decision-making, planning and purchasing processes. It works for well known corporations such as Amazon, Google, Microsoft and AT&T. This handbook is suitable for researchers and advanced-level students in computer science as a reference.
2,401 citations
TL;DR: A new neighborhood model with an improved prediction accuracy is introduced, which model neighborhood relations by minimizing a global cost function and makes both item-item and user-user implementations scale linearly with the size of the data.
Abstract: Recommender systems provide users with personalized suggestions for products or services. These systems often rely on collaborating filtering (CF), where past transactions are analyzed in order to establish connections between users and products. The most common approach to CF is based on neighborhood models, which originate from similarities between products or users. In this work we introduce a new neighborhood model with an improved prediction accuracy. Unlike previous approaches that are based on heuristic similarities, we model neighborhood relations by minimizing a global cost function. Further accuracy improvements are achieved by extending the model to exploit both explicit and implicit feedback by the users. Past models were limited by the need to compute all pairwise similarities between items or users, which grow quadratically with input size. In particular, this limitation vastly complicates adopting user similarity models, due to the typical large number of users. Our new model solves these limitations by factoring the neighborhood model, thus making both item-item and user-user implementations scale linearly with the size of the data. The methods are tested on the Netflix data, with encouraging results.
740 citations
25 Jun 2006
TL;DR: This work proposes a new algorithm, called BEETLE, for effective online learning that is computationally efficient while minimizing the amount of exploration, and takes a Bayesian model-based approach, framing RL as a partially observable Markov decision process.
Abstract: Reinforcement learning (RL) was originally proposed as a framework to allow agents to learn in an online fashion as they interact with their environment. Existing RL algorithms come short of achieving this goal because the amount of exploration required is often too costly and/or too time consuming for online learning. As a result, RL is mostly used for offline learning in simulated environments. We propose a new algorithm, called BEETLE, for effective online learning that is computationally efficient while minimizing the amount of exploration. We take a Bayesian model-based approach, framing RL as a partially observable Markov decision process. Our two main contributions are the analytical derivation that the optimal value function is the upper envelope of a set of multivariate polynomials, and an efficient point-based value iteration algorithm that exploits this simple parameterization.
297 citations
Proceedings Article•
26 Jun 2015TL;DR: In this paper, the authors consider reinforcement learning in parameterized Markov Decision Processes (MDPs) and derive a regret bound for priors over general parameter spaces, showing that the number of instants where suboptimal actions are chosen scales logarithmically with time, with high probability.
Abstract: We consider reinforcement learning in parameterized Markov Decision Processes (MDPs), where the parameterization may induce correlation across transition probabilities or rewards. Consequently, observing a particular state transition might yield useful information about other, unobserved, parts of the MDP. We present a version of Thompson sampling for parameterized reinforcement learning problems, and derive a frequentist regret bound for priors over general parameter spaces. The result shows that the number of instants where suboptimal actions are chosen scales logarithmically with time, with high probability. It holds for prior distributions that put significant probability near the true model, without any additional, specific closed-form structure such as conjugate or product-form priors. The constant factor in the logarithmic scaling encodes the information complexity of learning the MDP in terms of the Kullback-Leibler geometry of the parameter space.
62 citations
Patent•
PARC1
TL;DR: In this paper, the authors present a method for recommending activities to a user based on the user's personal profile and/or population prior information, thereby facilitating prediction of future activities for the user.
Abstract: One embodiment of the present invention provides a method for recommending activities to a user. During operation, the system determines an activity-type distribution based on the user's personal profile and/or population prior information, thereby facilitating prediction of future activities for the user. The system further searches for and receives one or more activities based on the activity-type distribution. The system then scores each received activity and recommends a number of activities to be performed by the user in the future and a number of corresponding venues, based on the activity-type distribution and the weight distribution.
44 citations