Topic modeling stands as a well-explored and foundational challenge in the text mining domain. Traditional topic schemes based on word co-occurrences, aim to expose the latent semantic structure embedded in a document corpus. Nevertheless, the inherent brevity of short texts introduces data sparsity, hindering the effectiveness of conventional topic models and yielding suboptimal outcomes for such text. Typically, short texts encompass a restricted number of topics, necessitating a grasp of relevant background knowledge for a comprehensive understanding of semantic content. Motivated by the observed information, this research introduces a novel Deep Auto encoder Graph Regularized Non-negative Matrix Factorization algorithm (DAGR-NMF) to uncover significant and meaningful topics within short document contents. The three main phases of proposed work are preprocessing, feature extraction and topic modeling. Initially, the data are preprocessed using natural language preprocessing tasks such as stop word removal, stemming and lemmatizing. Then, feature extraction is performed using hybrid Absolute Deviation Factors-Class Term Frequency (ADF-CTF) to capture the most relevant information from the text. Finally, topic modeling task is executed using proposed DAGR-NMF approach. Experimental findings demonstrate that the introduced DAGR-NMF model outperforms all other techniques by achieving NMI values of 0.852, 0.857, 0.793, and 0.831 on associated press, political blog datasets, 20NewsGroups, and News category dataset, respectively.

As the global demand for renewable energy increases, the photovoltaic (PV) industry, which is a vital component of clean energy, plays a crucial role in achieving energy transition and sustainable development goals. Consequently, the PV industry has grown rapidly in recent years, leading to an expanded and increasingly sophisticated industrial chain. However, the effective assessment of the development of the PV industry is complicated by the extensiveness of the PV industrial chain.

This paper presents a method for constructing a knowledge graph of the PV industry chain using enterprise bidding data, effectively coupling the product and supply networks. First, by leveraging relevant knowledge in the PV field, we employ two deep learning models, the BERT-BiLSTM-CRF model and an improved CasRel model, for entity and relationship extraction, respectively. Subsequently, entity-linking technology is applied to facilitate knowledge fusion. Finally, the Neo4j graph database is utilized for knowledge storage and graphical representation, comprehensively illustrating the technical process of constructing the PV industry chain knowledge graph.

The knowledge graph of the PV industry chain facilitates a timely understanding of the industry’s overall status and development trends while identifying bottlenecks and risks within the chain. Furthermore, it can aid enterprises in devising more effective risk management strategies and countermeasures, continuously optimizing the industry chain structure, and promoting the sustainable development of the industry.

In the realm of live streaming, timely and thoughtful responses play a pivotal role in enhancing user experience – a critical factor in the successful commercialization of live streaming. However, the inundation of user messages during live streaming poses a challenge for a streamer, necessitating a strategic approach to selective response. To determine which messages should be given priority for a response, exploring the underlying heterogeneity in user messages is vital. User messages do not contain explicit attitudes for segmentation but contain underlying intentions that can be inferred. Thus, in this study, we first employ a BERT-based text classifier to categorize user messages into two types: ritual messages, indicating socializing intentions, and functional messages, indicating information-seeking intentions. Based on the intention heterogeneity, we then analyze the relationship between two message types and user experience in live streaming and observe that functional messages tend to suggest the dissatisfaction of users. We further explore the influence of streamer responses on user experience with live streaming. Empirical results demonstrate that responses specifically targeted at functional messages have a significant positive impact on user experience, with this effect being moderated by streamer characteristics: gender and historical experience. These findings offer valuable guidance for streamers in tailoring their online response strategies effectively, and contribute to the literature on live streaming, user engagement behavior, and online responses, presenting actionable insights for providers and platforms of online service.

The controlling and synchronizing chaotic systems (CSs) are crucial aspects of engineering, with broad applications across various applied sciences, such as secure communications, nonlinear circuit design, biomedical engineering, and image processing. This paper deals with the complex problem of achieving finite-time projective synchronization for uncertain CSs with incommensurate non-integer orders using adaptive fuzzy sliding-mode control (AFSMC). Specifically, we focus on practical projective synchronization, introducing two novel control approaches that effectively mitigate the chattering phenomenon, a common issue in conventional sliding mode control. To achieve this, two innovative non-singular sliding surfaces with finite-time properties are formulated. This type of sliding surface enhances projective synchronization accuracy, response speed, and robustness. The adaptive fuzzy logic systems, known for their universal approximation capability, are employed to estimate continuous functional uncertainties. We rigorously analyzed the stability of both approaches using Lyapunov’s direct method. Extensive simulations confirm the effectiveness and benefits of our proposed methods. These methods significantly reduce or eliminate chattering and achieve practical projective synchronization in a finite time. This makes them well-suited for real-world applications in complex CSs.

The development of the Internet of Things (IoT) has rapidly progressed, revolutionizing numerous industries and transforming how we interact with technology. IoT relies on seamless connectivity between devices, allowing them to collect, share, and analyze data. As a result, compatibility and connectivity have become major concerns for customers adopting smart devices. Achieving compatibility in a smart device is a shared responsibility between the IoT platform provider and the smart device manufacturer. However, the impact of their respective efforts on smart device compatibility depends on investment efficiency and the service level. In this study, we develop a game-theoretical model to examine how investment efficiency affects the incentives of the upstream platform and the downstream manufacturer to exert compatibility efforts under the licensing pricing and revenue sharing models in a supply chain setting. Our findings indicate that, given a certain IoT platform service level, a manufacturer’s effort decreases as the relative weight decreases in the licensing pricing model. Additionally, when the platform’s effort cost is lower, the platform can invest more in compatibility, thereby improving the compatibility of smart devices. In such cases, the manufacturer can benefit from the platform’s incentives to exert compatibility efforts by reducing her efforts. We also identify two effects: the compatibility effect and the service value-added effect. When the relative weight of compatibility is small, the role of platform compatibility diminishes. Consequently, the platform cannot increase technology licensing fees to augment IoT service income, reducing profits. Our results also provide insights into how an IoT platform provider can strategically target a particular IoT service based on their cost characteristics and service level in smart device supply chains.