Large Language Models (LLMs) face challenges due to hallucination issues. Current solutions use retrieval-augmented generation (RAG), integrating LLMs with external knowledge to enhance answer accuracy. However, the misuse of irrelevant external knowledge can be misleading. In this paper, we propose a novel method called Retrieve-and-Discriminate Prompter (RD-P), which leverages knowledge graphs (KGs) for trustworthy RAG by synchronizing knowledge retrieval and discrimination in a unified model. Specifically, we train a prompter based on a pre-trained language model with shared parameters. It has two key modules: the retriever and the discriminator. The retriever identifies relevant reasoning paths in the KG, while the discriminator evaluates their credibility through "logical coverage calculation" and in turn instructs the retrieval process. Prompts are then constructed to guide LLMs in reasoning and answering questions using both retrieved and implicit knowledge. Experiments on knowledge-intensive question answering (QA) tasks demonstrate that our method significantly improves answer coverage rate while reducing the retrieval scale, achieving superior performance in complex KGQA tasks compared with state-of-the-art RAG methods at a low cost.

With the advent of 5G era, Internet-of-Things will become possible, and a large amount of correlative data, namely the streaming graph (SG), is continuously generated in various application fields, which poses a great challenge to analyze and make good use of such graph data. The balanced partition of a big graph, especially a dynamic SG, has always been basic research in graph theory, and there are many classical methods. However, this article explores a novel method, that is, an SG $k$ -way partitioning via a multiplayer repeated game so as to realize the equilibrium of the total number of nodes in different partitions and achieve high cohesion or low edge-cut in each partition. We regard the SG partitioning process during each time window as a multiplayer game process, treat $k$ target partitions as game players, and use all possible actions that players select new nodes as the strategy set. Some effective constraints are given to reduce the strategy space and accelerate the game process. Moreover, we define equilibrium degree (ED) and modularity degree (MD), which are used to design the utility function of game players. All game players finally reach the Nash equilibrium via multiplayer repeated games. At this time, each player corresponds to a strategy. Under the guidance of these strategies, the new nodes are added to each target partition, and then the goal of optimal partitioning is achieved. The above operations are carried out in each time window. The subsequent time windows can reuse the partitioning results of the previous time window in order to realize dynamic incremental partitioning for an SG. Extensive experiments indicate that while ensuring the partition quality of an SG, our proposed algorithm also greatly improves the speed of online partitioning. Even compared with the state-of-the-art SG partitioning algorithm PLDG, our algorithm still obtains better performance.

With the explosive growth of big data in various application areas, it is becoming very important for information management system to know the real-time growth change and the long-term increasing trend of big data. Thus, in this paper, we propose a big data growth scale prediction method based on deep learning intelligence. By analyzing the features of big data in various information systems, we summarize several basic data types which are ubiquitous in all kinds of information systems, and give the calculation method of each data growth scale time series. Considering the complex correlation among various types of data produced in information systems and the demand for data scale prediction, a deep learning prediction model, stacked autoencoders prediction (SAEP), is built. The previous time series prediction only considered one time series, but the SAEP can consider multiple time series simultaneously. We use data growth scale time series with different time interval as training sample sets to study features of big data in information systems, and then we can obtain the parameters of the SAEP model for the expected time interval. In order to solve the training problem of deep neural network, we also give a layer-wise training algorithm whose basic idea is to train all the hidden layers of the SAEP model layer by layer and then fine tune the parameters of the whole network. Repeated experiments show that the prediction performance of the SAEP model is stable and obviously better than the traditional exponential regression prediction method.