13 results about "Recommendation quality" patented technology

The invention provides an intelligent supplier recommendation method and system, and relates to the technical field of supply chain management, and the method comprises the steps: 1, mapping supplier evaluation index data to a multi-dimensional space, generating an independent space coordinate corresponding to each supplier, and forming a space coordinate set; and step 2, calculating a spatial distribution core coordinate based on the spatial coordinate set, establishing two orthogonal reference direction lines by taking the core coordinate as an original point, and delimiting an evaluation range area according to an opening angle formed by the reference direction lines. According to the method, the suppliers can be recommended by objectively integrating the multi-dimensional indexes, and dynamic updating can be performed to guarantee the recommendation quality.

The invention discloses a performance self-improving large model recommendation system and method, and the system comprises an initialization module which is used for pre-training a large language model through supervision and fine tuning, and generating an initial recommendation model; the self-optimization module comprises three iteratively executed sub-modules; the sample selection sub-module is used for screening historical data samples of which the information amount is higher than a threshold value on the basis of comparison between the model prediction probability and the preset threshold value; the response fusion sub-module is used for generating K candidate responses for a selected sample and generating a preference data set based on model evaluation; and the DPO optimization sub-module is used for updating model parameters by utilizing the improved DPO loss function and generating an optimized recommendation model. According to the method, the dependence on static preference data in the prior art is broken, the recommendation quality and robustness are improved, and adaptive optimization is realized.

The invention discloses an adversarial simulation learning method for relieving popularity deviation in multiple behaviors, and solves the problem of popularity deviation under multi-behavior recommendation by adopting a multi-task joint optimization form and combining technologies such as an adversarial network and simulation learning. According to the recommendation method, a post-order behavior module is used as a generator, a preorder behavior module is used as a discriminator, a simulation adversarial learning framework is constructed, the generator can generate tail project embedding simulating preorder behavior characteristics, project embedding of preorder behaviors is enhanced through a gating fusion strategy, the recommendation quality is guaranteed, and meanwhile the recommendation efficiency is improved. And finally, the popularity deviation problem is effectively relieved.

The invention discloses a personalized recommendation method based on differential privacy and federated learning, which comprises the following steps of: firstly, collecting user behavior data and carrying out preprocessing and behavior modeling on the user behavior data, and then adopting a dynamic privacy budget allocation and differential privacy noise injection mechanism in a local training stage of federated learning; meanwhile, an adaptive gating layer is introduced to cut low-importance parameters, then data modeling and personalized model training are carried out based on a Gaussian mixture model, finally, global aggregation of federated learning is carried out by adopting a personalized model aggregation strategy, and a global model is dynamically updated according to data distribution of different clients. And the local personalized recommendation effect is ensured. According to the method, the problems of privacy protection, overlarge calculation and communication overhead, data heterogeneity and the like in a personalized recommendation system are effectively solved, the model training efficiency and the recommendation quality are improved, and the method is suitable for multiple fields of e-commerce, social platforms, video recommendation and the like and has wide application value.

The invention provides an AI platform data recommendation method and device and computer equipment, and relates to the technical field of data processing, and the method comprises the steps: obtaining a historical consultation dialogue record set of a target user, carrying out progressive multi-round dialogue iterative memory analysis, and determining the understanding category of the target user; obtaining a real-time active reasoning dialogue sequence; when a backtracking detection module is called to carry out recommended data completeness detection on the real-time active reasoning dialogue sequence, a recommended data completeness detection result is obtained; and triggering a recommendation output instruction according to the recommendation data completeness coefficient, and pushing recommendation data based on the recommendation output instruction. The data recommendation method and device have the advantages that the technical problems that in the data recommendation process in the prior art, in-depth interaction with the user cannot be dynamically conducted, initiative questioning in accordance with the actual situation of the user cannot be achieved, the recommendation quality is low, and the real-time requirement of the user cannot be met easily are solved, and the technical effect of improving the data recommendation quality is achieved.

The invention discloses a personalized robust recommendation defense method based on a dynamic reward mechanism, and the method comprises the steps: obtaining a real world data set related to a recommendation system, carrying out the cleaning of the data set, filtering invalid interaction data, retaining an effective user-article interaction pair, carrying out the normalization of scoring data, and dividing the scoring data into a training set, a verification set and a test set; constructing a robustness and fairness dual-objective influence evaluation mechanism, and quantifying an influence value of each user-article interaction sample on the model; the disturbance resisting amplitude is dynamically adjusted according to different user behavior characteristics; designing a reward mechanism integrating embedded interaction, behavior similarity and optimal batch, and guiding the model to dynamically balance defense and recommendation quality; and based on the filtered data set, training a model in combination with personalized adversarial disturbance and a multi-dimensional reward function. According to the method, pollution of poisoning samples to the model is effectively reduced, the adaptability of the model to heterogeneous user behaviors is improved, the attack resistance of the system is remarkably enhanced, and the method has wide applicability and practicability.

The invention discloses a computer network security situation intelligent prediction method and system based on big data, and particularly relates to the technical field of data analysis. The method comprises the following steps: acquiring CPU / GPU utilization rate, memory consumption, I / O throughput and user feedback data of a server in real time, calculating a ratio of recommendation request handling capacity per second to a system bearable threshold value, monitoring abnormal fluctuation of the user feedback data, constructing a data prediction model, calculating whether calculation resources of a recommendation system are overloaded or not, and sending out early warning before overload occurs. According to the method, optimization processing is automatically executed, the implementation effect of an optimization strategy is continuously monitored, the change trend of computing resources is analyzed, and if abnormity is detected, load balance and a recommendation strategy are dynamically adjusted, so that efficient utilization of the computing resources and stability of recommendation quality are ensured, and through intelligent analysis and decision making, the reliability of a recommendation system is improved; and the overload risk of computing resources is reduced, the user experience is optimized, and the adaptive capacity of the system in a high-concurrency scene is enhanced.

There are proposed methods, devices, and computer program products for recommendation. In the method, a first group of state data associated with a first group of entities in a recommendation system is obtained. A second group of state data associated with a second group of entities in the recommendation system is obtained. A first and a second group of objects that are to be recommended to the first and second groups of entities are determined by a machine learning model based on the first and second groups of state data respectively, so that an association relationship between a first and a second group of state transitions of the first and second groups of entities meets a predetermined condition. With these implementations, recommendation quality may be continuously improved and the disparity in retention may be minimized across groups of entities over time while enforcing retention fairness in a long term.

The invention relates to a social recommendation method integrating adaptive graph features and comparison optimization, and discloses a novel social recommendation method integrating adaptive graph features and comparison optimization. By fusing information of a user-article interaction graph and a social graph, the method is innovatively designed for a cold start problem and a data sparsity challenge in social recommendation. The core technical features comprise: (1) a graph fusion feature network structure, dynamically calculating feature weights between nodes and neighbors, effectively fusing features of different graphs, and selectively retaining important features; and (2) a self-adaptive contrast learning framework is adopted, and positive and negative sample pairs are constructed, so that the relevance between the target behavior and the auxiliary behavior can be enhanced, and the sparsity of the target behavior can be relieved. The main purpose of the model is to provide personalized recommendation service for the user, especially in cold start and data sparse scenes, the performance is excellent, the recommendation quality is significantly improved, and the method has high practicability and innovativeness.

The invention discloses a commodity recommendation method based on big data, and the method comprises the steps: collecting data, recognizing the credibility of the data, carrying out the statement analysis of the data based on a big model, and generating an information credibility score; calculating attribute similarity and semantic similarity, calculating basic similarity, and correcting the basic similarity to obtain final similarity; a weight mapping table is set, the system performs automatic query according to commodity categories, obtains an optimal weight combination and performs descending sort according to final similarity to obtain a similar commodity recommendation table, and a large model is used for generating an explanatory reason for a recommendation result in the similar commodity recommendation table; when similar commodities are insufficient or are new commodities, a popularity substitution strategy is used, mixed calculation of attribute similarity and semantic similarity is adopted, recommendation accuracy is improved, information credibility scores are introduced to serve as weight factors, similarity calculation is more accurate, interference of low-credibility commodities on recommendation results is reduced, and recommendation quality is improved.

The application discloses a method, device, storage medium and electronic equipment for generating a recommendation list. The method comprises the following steps: constructing a project structure diagram containing a first object and a target project, determining a connection matrix between the first object and the target project from the project structure diagram; calculating the similarity between the first object and other objects according to the connection matrix; obtaining a preset similarity, screening other objects based on the preset similarity and the similarity, and obtaining a second object set similar to the first object; connecting the first object and each second object in the second object set in a graph to obtain a relationship graph of the first object and each second object; obtaining a relationship matrix of the object and the target project, and generating a recommendation list according to the relationship graph and the relationship matrix. The application solves the technical problems of inaccurate recommendation results, poor recommendation quality and cold start caused by ignoring the edge information of users or projects in the process of performing a recommendation task in the related art.

The application provides a supplier intelligent recommendation method and system, and relates to the technical field of supply chain management. The method comprises the following steps: in step 1, supplier evaluation index data is mapped to a multidimensional space, independent space coordinates corresponding to each supplier are generated, and a space coordinate set is formed; in step 2, based on the space coordinate set, a space distribution core coordinate is calculated, two orthogonal reference direction lines are established with the core coordinate as the origin, and an evaluation range area is delimited according to the opening angle formed by the reference direction lines. The application can objectively and comprehensively recommend suppliers in multiple dimensions, and can be dynamically updated to guarantee the recommendation quality.

The application relates to a heterogeneous collaborative filtering method for multi-behavior recommendation, belongs to the multi-behavior recommendation field, and is based on a heterogeneous collaborative filtering model HCFMR for multi-behavior recommendation and comprises the following steps: S1, obtaining user and item interaction features of a specific behavior by using an improved lightweight graph convolutional network; S2, obtaining multi-behavior dependent semantics of a user bottom layer by using a multi-head attention network; S3, setting a weight of a specific behavior for each user, and the weight is used for distinguishing the contribution degrees of different behaviors; and S4, obtaining high-order embedding representation of a node by aggregating various convolutional layers, obtaining a user preference, and then performing multi-behavior recommendation. The application reduces model complexity, enriches user and item embedding representation, improves recommendation quality, and improves the explainability of the model.