Generalized Category Discovery (GCD) aims to identify both known and unknown categories from unlabeled data using limited labeled samples from known classes. The key challenges lie in the scarcity of supervision signals for unknown categories and the difficulty of modeling relationships among samples. Existing methods that rely solely on clustering uncertainty often result in imperfect hard negative selection, while their use of nearest-neighbor structures hinders the effective utilization of Large Language Model (LLM) annotations for obtaining high-quality supervision. We propose a dynamic optimization model (LIGD) that leverages diffusion graphs and LLM annotations to address these issues. By leveraging the semantic-correlation graph, the method achieves two key capabilities: it selects both hard negatives and unlabeled central samples likely to represent novel categories as high-value samples. In addition, the graph enables effective label propagation through its connected subgraph, significantly reducing computational costs while enhancing the accuracy of category discovery. To further enhance annotation quality, we introduce a two-stage prompting strategy that queries the LLM twice to accurately assign selected samples to either existing or novel categories. The entire process will be repeated iteratively until convergence to update the graph structure and node representations in the graph. Experiments on three GCD datasets demonstrate the significant superiority of LIGD. Most notably, in the challenging scenario where only 25% of categories are labeled, the model achieves substantial improvements while reducing the number of LLM queries by 50%. Code and data are available at https://github.com/wdmmxlbt/LIGD.

Multimodal recommender systems utilize various types of information to model user preferences and item features, helping users discover items aligned with their interests. The integration of multimodal information mitigates the inherent challenges in recommender systems, e.g., the data sparsity problem and cold-start issues. However, it simultaneously magnifies certain risks from multimodal information inputs, such as information adjustment risk and inherent noise risk. These risks pose crucial challenges to the robustness of recommendation models. In this paper, we analyze multimodal recommender systems from the novel perspective of flat local minima and propose a concise yet effective gradient strategy called Mirror Gradient (MG). This strategy can implicitly enhance the model's robustness during the optimization process, mitigating instability risks arising from multimodal information inputs. We also provide strong theoretical evidence and conduct extensive empirical experiments to show the superiority of MG across various multimodal recommendation models and benchmarks. Furthermore, we find that the proposed MG can complement existing robust training methods and be easily extended to diverse advanced recommendation models, making it a promising new and fundamental paradigm for training multimodal recommender systems. The code is released at https://github.com/Qrange-group/Mirror-Gradient.

Abstract Scientific novelty drives the efforts to invent new vaccines and solutions during the pandemic. First‐time collaboration and international collaboration are two pivotal channels to expand teams' search activities for a broader scope of resources required to address the global challenge, which might facilitate the generation of novel ideas. Our analysis of 98,981 coronavirus papers suggests that scientific novelty measured by the BioBERT model that is pretrained on 29 million PubMed articles, and first‐time collaboration increased after the outbreak of COVID‐19, and international collaboration witnessed a sudden decrease. During COVID‐19, papers with more first‐time collaboration were found to be more novel and international collaboration did not hamper novelty as it had done in the normal periods. The findings suggest the necessity of reaching out for distant resources and the importance of maintaining a collaborative scientific community beyond nationalism during a pandemic.

Industrial big data is the key to realize the vision of smart factories. This research aims to identify and explore potential barriers that prevent organizations from deploying industrial big data solutions in the development of smart factories through a socio-technical perspective. The research follows an inductive qualitative approach. Twenty-seven semi-structured interviews were conducted with the CEO, smart factory manager, IT managers, departmental heads, and IS consultants in the selected case company. The interview data were analyzed using a thematic analysis method. Derived from a thematic analysis, six sets of barriers including technical, data, technical support, organization, individual, and social issues were identified, as well as the relationships between them. An empirical framework was developed to highlight the relationship between these barriers. This study contributes to the knowledge of industrial big data in general and provides constructive insight into industrial big data implementation in smart factory development particularly.

Attributed network embedding (ANE) attempts to represent a network in short code, while retaining information about node topological structures and node attributes. A node's feature and topological structure information could be divided into different local aspects, while in many cases, not all the information but part of the information contained in several local aspects determine the relations among different nodes. Most of the existing works barely concern and identify the aspect influence from network embedding to our knowledge. We attempt to use local embeddings to represent local aspect information and propose InfomaxANE which encodes both global and local embeddings from the perspective of mutual information. The local aspect embeddings are forced to learn and extract different aspect information from nodes' features and topological structures by using orthogonal constraint. A theoretical analysis is also provided to further confirm its correctness and rationality. Besides, to provide complete and refined information for local encoders, we also optimize feature aggregation in SAGE with different structures: feature similarities are concerned and aggregator is seperated from encoder. InfomaxANE is evaluated on both node clustering and node classification tasks (including both transductive and inductive settings) with several benchmark datasets, the results show the outperformance of InfomaxANE over competitive baselines. We also verify the significance of each module in our proposed InfomaxANE in the additional experiment.

Information from microblogs is gaining increasing attention from researchers interested in analyzing fluctuations in stock markets. Behavioral financial theory draws on social psychology to explain some of the irrational behaviors associated with financial decisions to help explain some of the fluctuations. In this study we argue that social media users who demonstrate an interest in finance can offer insights into ways in which irrational behaviors may affect a stock market. To test this, we analyzed all the data collected over a 3‐month period in 2011 from Tencent Weibo (one of the largest microblogging websites in China). We designed a social influence (SI)‐based Tencent finance‐related moods model to simulate investors' irrational behaviors, and designed a Tencent Moods‐based Stock Trend Analysis (TM_STA) model to detect correlations between Tencent moods and the Hushen‐300 index (one of the most important financial indexes in China). Experimental results show that the proposed method can help explain the data fluctuation. The findings support the existing behavioral financial theory, and can help to understand short‐term rises and falls in a stock market. We use behavioral financial theory to further explain our findings, and to propose a trading model to verify the proposed model.

With the information overload of user‐generated content in microblogging, users find it extremely challenging to browse and find valuable information in their first attempt. In this paper we propose a microblogging recommendation algorithm, TSI‐MR (Topic‐Level Social Influence‐based Microblogging Recommendation), which can significantly improve users' microblogging experiences. The main innovation of this proposed algorithm is that we consider social influences and their indirect structural relationships, which are largely based on social status theory, from the topic level. The primary advantage of this approach is that it can build an accurate description of latent relationships between two users with weak connections, which can improve the performance of the model; furthermore, it can solve sparsity problems of training data to a certain extent. The realization of the model is mainly based on Factor Graph. We also applied a distributed strategy to further improve the efficiency of the model. Finally, we use data from Tencent Weibo, one of the most popular microblogging services in China, to evaluate our methods. The results show that incorporating social influence can improve microblogging performance considerably, and outperform the baseline methods.

Text mining has been widely used in multiple types of user‐generated data to infer user opinion, but its application to microblogging is difficult because text messages are short and noisy, providing limited information about user opinion. Given that microblogging users communicate with each other to form a social network, we hypothesize that user opinion is influenced by its neighbors in the network. In this paper, we infer user opinion on a topic by combining two factors: the user's historical opinion about relevant topics and opinion influence from his/her neighbors. We thus build a topic‐level opinion influence model (TOIM) by integrating both topic factor and opinion influence factor into a unified probabilistic model. We evaluate our model in one of the largest microblogging sites in China, Tencent Weibo, and the experiments show that TOIM outperforms baseline methods in opinion inference accuracy. Moreover, incorporating indirect influence further improves inference recall and f1‐measure. Finally, we demonstrate some useful applications of TOIM in analyzing users' behaviors in Tencent Weibo.

This paper proposes a Topic-Level Opinion Influence Model (TOIM) that simultaneously incorporates topic factor, user opinions and social influence in a unified probabilistic model with two stages learning processes. In the first stage, topic factor and user influence are integrated to generate users' influential relationship based on different topics; in the second stage, users' historical messages and social interaction records are leveraged by TOIM to construct their historical opinions and neighbors' opinion influence through a statistical learning process, which can be further utilized to predict users' future opinions on some specific topics. We evaluate our TOIM on a large-scaled dataset from Tencent Weibo, one of the largest microbloggings website in China. The experimental results show that TOIM can better predict users' opinion than other baseline methods.

Abstract This article investigates the dynamic features of social tagging vocabularies in Delicious, Flickr, and YouTube from 2003 to 2008. Three algorithms are designed to study the macro‐ and micro‐tag growth as well as the dynamics of taggers' activities, respectively. Moreover, we propose a Tagger Tag Resource Latent Dirichlet Allocation (TTR‐LDA) model to explore the evolution of topics emerging from those social vocabularies. Our results show that (a) at the macro level, tag growth in all the three tagging systems obeys power law distribution with exponents lower than 1; at the micro level, the tag growth of popular resources in all three tagging systems follows a similar power law distribution; (b) the exponents of tag growth vary in different evolving stages of resources; (c) the growth of number of taggers associated with different popular resources presents a feature of convergence over time; (d) the active level of taggers has a positive correlation with the macro‐tag growth of different tagging systems; and (e) some topics evolve into several subtopics over time while others experience relatively stable stages in which their contents do not change much, and certain groups of taggers continue their interests in them.

The presence of social networks in complex systems has made networks and community structure a focal point of study in many domains. Previous studies have focused on the structural emergence and growth of communities and on the topics displayed within the network. However, few scholars have closely examined the relationship between the thematic and structural properties of networks. Therefore, this article proposes the Tagger Tag Resource-Latent Dirichlet Allocation-Community model (TTR-LDA-Community model), which combines the Latent Dirichlet Allocation (LDA) model with the Girvan-Newman community detection algorithm through an inference mechanism. Using social tagging data from Delicious, this article demonstrates the clustering of active taggers into communities, the topic distributions within communities, and the ranking of taggers, tags, and resources within these communities. The data analysis evaluates patterns in community structure and topical affiliations diachronically. The article evaluates the effectiveness of community detection and the inference mechanism embedded in the model and finds that the TTR-LDA-Community model outperforms other traditional models in tag prediction. This has implications for scholars in domains interested in community detection, profiling, and recommender systems.

This work identifies changes in dominant topics in library and information science (LIS) over time, by analyzing the 3,121 doctoral dissertations completed between 1930 and 2009 at North American Library and Information Science programs. The authors utilize latent Dirichlet allocation (LDA) to identify latent topics diachronically and to identify representative dissertations of those topics. The findings indicate that the main topics in LIS have changed substantially from those in the initial period (1930–1969) to the present (2000–2009). However, some themes occurred in multiple periods, representing core areas of the field: library history occurred in the first two periods; citation analysis in the second and third periods; and information-seeking behavior in the fourth and last period. Two topics occurred in three of the five periods: information retrieval and information use. One of the notable changes in the topics was the diminishing use of the word library (and related terms). This has implications for the provision of doctoral education in LIS. This work is compared to other earlier analyses and provides validation for the use of LDA in topic analysis of a discipline.

This article investigates the dynamic features of social tagging vocabularies in Delicious, Flickr and YouTube from 2003 to 2008. It analyzes the evolution of the usage of the most popular tags in each of these three social networks. We find that for different tagging systems, the dynamic features reflect different cognitive processes. At the macro level, the tag growth obeys power-law distribution for all three tagging systems with exponents lower than one. At the micro level, the tag growth of popular resources in all three tagging systems follows a similar power-law distribution. Moreover, we find that the exponents of tag growth varied in different evolving stages of popular individual resources.

Exploring community is fundamental for uncovering the connections between structure and function of complex networks and for practical applications in many disciplines such as biology and sociology. In this paper, we propose a TTR-LDA-Community model which combines the Latent Dirichlet Allocation model (LDA) and the Girvan-Newman community detection algorithm with an inference mechanism. The model is then applied to data from Delicious, a popular social tagging system, over the time period of 2005-2008. Our results show that 1) users in the same community tend to be interested in similar set of topics in all time periods; and 2) topics may divide into several sub-topics and scatter into different communities over time. We evaluate the effectiveness of our model and show that the TTR-LDA-Community model is meaningful for understanding communities and outperforms TTR-LDA and LDA models in tag prediction.