Sentiment analysis of online Chinese buzzwords (OCBs) is important for healthy development of platforms, such as games and social networking, which can avoid transmission of negative emotions through prediction of users’ sentiment tendencies. Buzzwords have the characteristics of varying text length, irregular wording, ignoring syntactic and grammatical requirements, no complete semantic structure, and no obvious sentiment features. This results in interference and challenges to the sentiment analysis of such texts. Sentiment analysis also requires capturing effective sentiment features from deeper contextual information. To solve the above problems, we propose a deep learning model combining BERT and BiLSTM. The goal is to generate dynamic representations of OCB vectors in downstream tasks by fine-tuning the BERT model and to capture the rich information of the text at the embedding layer to solve the problem of static representations of word vectors. The generated word vectors are then transferred to the BiLSTM model for feature extraction to obtain the local and global semantic features of the text while highlighting the text sentiment polarity for sentiment classification. The experimental results show that the model works well in terms of the comprehensive evaluation index F1. Our model also has important significance and research value for sentiment analysis of irregular texts, such as OCBs.

Knowledge graph embedding (KGE) can benefit a variety of downstream tasks, such as link prediction and relation extraction, and has therefore quickly gained much attention. However, most conventional embedding models assume that all triple facts share the same confidence without any noise, which is inappropriate. In fact, many noises and conflicts can be brought into a knowledge graph (KG) because of both the automatic construction process and data quality problems. Fortunately, the novel confidence-aware knowledge representation learning (CKRL) framework was proposed, to incorporate triple confidence into translation-based models for KGE. Though effective at detecting noises, with uniform negative sampling methods, and a harsh triple quality function, CKRL could easily cause zero loss problems and false detection issues. To address these problems, we introduce the concept of negative triple confidence and propose a confidence-aware negative sampling method to support the training of CKRL in noisy KGs. We evaluate our model on the knowledge graph completion task. Experimental results demonstrate that the idea of introducing negative triple confidence can greatly facilitate performance improvement in this task, which confirms the capability of our model in noisy knowledge representation learning (NKRL).

Graph edge partitioning, which is essential for the efficiency of distributed graph computation systems, divides a graph into several balanced partitions within a given size to minimize the number of vertices to be cut. Existing graph partitioning models can be classified into two categories: offline and streaming graph partitioning models. The former requires global graph information during the partitioning, which is expensive in terms of time and memory for large-scale graphs. The latter creates partitions based solely on the received graph information. However, the streaming model may result in a lower partitioning quality compared with the offline model. Therefore, this study introduces a Local Graph Edge Partitioning model, which considers only the local information (i.e., a portion of a graph instead of the entire graph) during the partitioning. Considering only the local graph information is meaningful because acquiring complete information for large-scale graphs is expensive. Based on the Local Graph Edge Partitioning model, two local graph edge partitioning algorithms—Two-stage Local Partitioning and Adaptive Local Partitioning—are given. Experimental results obtained on 14 real-world graphs demonstrate that the proposed algorithms outperform rival algorithms in most tested cases. Furthermore, the proposed algorithms are proven to significantly improve the efficiency of the real graph computation system GraphX.

Knowledge graph completion (KGC) involves inferring missing entities or relationships within a knowledge graph, playing a crucial role across various domains, including intelligent question answering, recommendation systems, and dialogue systems. Traditional knowledge graph embedding (KGE) methods have proven effective in utilizing structured data and relationships. However, these methods often overlook the vast amounts of unstructured data and the complex reasoning capabilities required to handle ambiguous queries or rare entities. Recently, the rapid development of large language models (LLMs) has demonstrated exceptional potential in text comprehension and contextual reasoning, offering new prospects for KGC tasks. By using traditional KGE to capture the structural information of entities and relations to generate candidate entities and then reranking them with a generative LLM, the output of the LLM can be constrained to improve reliability. Despite this, new challenges, such as omissions and incorrect responses, arise during the ranking process. To address these issues, a knowledge-guided LLM reasoning for knowledge graph completion (KLR-KGC) framework is proposed. This model retrieves two types of knowledge from the knowledge graph—analogical knowledge and subgraph knowledge—to enhance the LLM’s logical reasoning ability for specific tasks while injecting relevant additional knowledge. By integrating a chain-of-thought (CoT) prompting strategy, the model guides the LLM to filter and rerank candidate entities, constraining its output to reduce omissions and incorrect responses. The framework aims to learn and uncover the latent correspondences between entities, guiding the LLM to make reasonable inferences based on supplementary knowledge for more accurate predictions. The experimental results demonstrate that on the FB15k-237 dataset, KLR-KGC outperformed the entity generation model (CompGCN), achieving a 4.8% improvement in MRR and a 5.8% improvement in Hits@1.

Text clustering is the task of grouping text data based on similarity, and it holds particular importance in the medical field. sIn healthcare, medical data clustering is a highly active and effective research area. It not only provides strong support for making correct medical decisions from medical datasets but also aids in patient record management and medical information retrieval. With the development of the healthcare industry, a large amount of medical data is being generated, and traditional medical data clustering faces significant challenges. Many existing text clustering algorithms are primarily based on the bag-of-words model, which has issues such as high dimensionality, sparsity, and the neglect of word positions and context. Pre-trained models are a deep learning-based approach that treats text as a sequence to accurately capture word positions and context information. Moreover, compared to traditional K-means and fuzzy C-means clustering models, deep learning-based clustering algorithms are better at handling high-dimensional, complex, and nonlinear data. In particular, clustering algorithms based on autoencoders can learn data representations and clustering information, effectively reducing noise interference and errors during the clustering process. This paper combines pre-trained language models with deep embedding clustering models. Experimental results demonstrate that our model performs exceptionally well on four public datasets, outperforming most existing text clustering algorithms, and can be applied to medical data clustering.

The superior performance of large-scale pre-trained models, such as Bidirectional Encoder Representations from Transformers (BERT) and Generative Pre-trained Transformer (GPT), has received increasing attention in both academic and industrial research and has become one of the current research hotspots. A pre-trained model refers to a model trained on large-scale unlabeled data, whose purpose is to learn general language representation or features for fine-tuning or transfer learning in subsequent tasks. After pre-training is complete, a small amount of labeled data can be used to fine-tune the model for a specific task or domain. This two-stage method of “pre-training+fine-tuning” has achieved advanced results in natural language processing (NLP) tasks. Despite widespread adoption, existing fixed fine-tuning schemes that adapt well to one NLP task may perform inconsistently on other NLP tasks given that different tasks have different latent semantic structures. In this article, we explore the effectiveness of automatic fine-tuning pattern search for layer-wise learning rates from an evolutionary optimization perspective. Our goal is to use evolutionary algorithms to search for better task-dependent fine-tuning patterns for specific NLP tasks than typical fixed fine-tuning patterns. Experimental results on two real-world language benchmarks and three advanced pre-training language models show the effectiveness and generality of the proposed framework.