Explore the vast range of titles available.

The fundamental goal of artificial intelligence (AI) is to mimic the core cognitive activities of human. Despite tremendous success in the AI research, most of existing methods have only single-cognitive ability. To overcome this limitation and take a solid step towards artificial general intelligence (AGI), we develop a foundation model pre-trained with huge multimodal data, which can be quickly adapted for various downstream cognitive tasks. To achieve this goal, we propose to pre-train our foundation model by self-supervised learning with weak semantic correlation data crawled from the Internet and show that promising results can be obtained on a wide range of downstream tasks. Particularly, with the developed model-interpretability tools, we demonstrate that strong imagination ability is now possessed by our foundation model. We believe that our work makes a transformative stride towards AGI, from our common practice of “weak or narrow AI” to that of “strong or generalized AI”. Artificial intelligence approaches inspired by human cognitive function have usually single learned ability. The authors propose a multimodal foundation model that demonstrates the cross-domain learning and adaptation for broad range of downstream cognitive tasks.

It appears that nationalism has been on the rise in China in recent years, particularly among online communities. Scholars agree that the Chinese government is facing pressure from online nationalistic and pro-democracy forces; however, it is believed that of the two, nationalistic views are the more dominant. Online nationalism is believed to have pushed the Chinese government to be more aggressive in diplomacy. This study challenges this conventional wisdom by finding that online political discourse is not dominated by nationalistic views, but rather by anti-regime sentiments. Even when there is an outpouring of nationalist sentiment, it may be accompanied by pro-democracy views that criticize the government. By analysing more than 6,000 tweets from 146 Chinese opinion leaders on Weibo, and by decomposing nationalistic discussion by specific topic, this study shows that rather than being monolithically xenophobic, nationalists may have differing sets of views regarding China's supposed rivals. Rather than being supportive of the regime, nationalists may incorporate liberal values to challenge the government. Nonetheless, this liberal dominance appears to provoke a backlash of nationalism among certain groups. 近年来，民族主义情绪看似在中国网络空间节节走高。学者们认为互联网给中国政府带来了民族主义的压力，并促使政府在外交上日趋铁腕。然而，本研究发现，民族主义情绪并非网络政治话语的主导部分。即便在其泛滥之时，民族主义情绪也可能与自由主义意见相结合，并对政府提出批评。本研究对 146 名微博意见领袖共计 6000 余条微博进行了内容分析，并将民族主义讨论拆解为不同的话题进行比较。我们发现，民族主义者并不是站在统一战线的仇外主义者，反而对所谓中国的敌手抱有不同的好恶; 同时，民族主义者并不是无条件拥护政府，反而有可能吸纳自由派的意见来质疑政府的权威。然而，强势的自由主义却也在一定程度上引发反噬，在某些网络群体中激发了爱国主义和民族主义情绪。

To develop a knowledge-aware recommender system, a key issue is how to obtain rich and structured knowledge base (KB) information for recommender system (RS) items. Existing data sets or methods either use side information from original RSs (containing very few kinds of useful information) or utilize a private KB. In this paper, we present , a data set linking KB information for RSs. It has linked three widely used RS data sets with two popular KBs, namely Freebase and YAGO. Based on our linked data set, we first preform qualitative analysis experiments, and then we discuss the effect of two important factors (i.e., popularity and recency) on whether a RS item can be linked to a KB entity. Finally, we compare several knowledge-aware recommendation algorithms on our linked data set.

Key Smart City applications such as traffic management and public security rely heavily on the intelligent processing of video and image data, often in the form of visual retrieval tasks, such as person Re-IDentification (ReID) and vehicle re-identification. For these tasks, Deep Neural Networks (DNNs) have been the dominant solution for the past decade, for their remarkable ability in learning discriminative features from images to boost retrieval performance. However, it is been discovered that DNNs are broadly vulnerable to maliciously constructed adversarial examples. By adding small perturbations to a query image, the returned retrieval results will be completely dissimilar from the query image. This poses serious challenges to vital systems in Smart City applications that depend on the DNN-based visual retrieval technology, as in the physical world, simple camouflage can be added on the subject (a few patches on the body or car), and turn the subject completely untrackable by person or vehicle Re-ID systems. To demonstrate the potential of such threats, this paper proposes a novel adversarial patch generative adversarial network (AP-GAN) to generate adversarial patches instead of modifying the entire image, which also causes the DNNs-based image retrieval models to return incorrect results. AP-GAN is trained in an unsupervised way that requires only a small amount of unlabeled data for training. Once trained, it produces query-specific perturbations for query images to form adversarial queries. Extensive experiments show that the AP-GAN achieves excellent attacking performance with various application scenarios that are based on deep features, including image retrieval, person ReID and vehicle ReID. The results of this study provide a warning that when deploying a DNNs-based image retrieval system, its security and robustness needs to be thoroughly considered.

Web search provides a promising way for people to obtain information and has been extensively studied. With the surge of deep learning and large-scale pre-training techniques, various neural information retrieval models are proposed, and they have demonstrated the power for improving search (especially, the ranking) quality. All these existing search methods follow a common paradigm, i.e., index-retrieve-rerank, where they first build an index of all documents based on document terms (i.e., sparse inverted index) or representation vectors (i.e., dense vector index), then retrieve and rerank retrieved documents based on the similarity between the query and documents via ranking models. In this paper, we explore a new paradigm of information retrieval without an explicit index but only with a pre-trained model. Instead, all of the knowledge of the documents is encoded into model parameters, which can be regarded as a differentiable indexer and optimized in an end-to-end manner. Specifically, we propose a pre-trained model-based information retrieval (IR) system called DynamicRetriever, which directly returns document identifiers for a given query. Under such a framework, we implement two variants to explore how to train the model from scratch and how to combine the advantages of dense retrieval models. Compared with existing search methods, the model-based IR system parameterizes the traditional static index with a pre-training model, which converts the document semantic mapping into a dynamic and updatable process. Extensive experiments conducted on the public search benchmark Microsoft machine reading comprehension (MS MARCO) verify the effectiveness and potential of our proposed new paradigm for information retrieval.

In many e-commerce Web sites, product recommendation is essential to improve user experience and boost sales. Most existing product recommender systems rely on historical transaction records or Web-site-browsing history of consumers in order to accurately predict online users’ preferences for product recommendation. As such, they are constrained by limited information available on specific e-commerce Web sites. With the prolific use of social media platforms, it now becomes possible to extract product demographics from online product reviews and social networks built from microblogs. Moreover, users’ public profiles available on social media often reveal their demographic attributes such as age, gender, and education. In this paper, we propose to leverage the demographic information of both products and users extracted from social media for product recommendation. In specific, we frame recommendation as a learning to rank problem which takes as input the features derived from both product and user demographics. An ensemble method based on the gradient-boosting regression trees is extended to make it suitable for our recommendation task. We have conducted extensive experiments to obtain both quantitative and qualitative evaluation results. Moreover, we have also conducted a user study to gauge the performance of our proposed recommender system in a real-world deployment. All the results show that our system is more effective in generating recommendation results better matching users’ preferences than the competitive baselines.

SimRank is a popular measurement for evaluating the node-to-node similarities based on the graph topology. In recent years, single-source and top-k SimRank queries have received increasing attention due to their applications in web mining, social network analysis, and spam detection. However, a fundamental obstacle in studying SimRank has been the lack of ground truths. The only exact algorithm, Power Method, is computationally infeasible on graphs with more than 106 nodes. Consequently, no existing work has evaluated the actual accuracy of various single-source and top-k SimRank algorithms on large real-world graphs. In this paper, we present ExactSim, the first algorithm that computes the exact single-source and top-k SimRank results on large graphs. This algorithm produces ground truths with precision up to 7 decimal places with high probability. With the ground truths computed by ExactSim, we present the first experimental study of the accuracy/cost trade-offs of existing approximate SimRank algorithms on large real-world graphs and synthetic graphs. Finally, we use the ground truths to exploit various properties of SimRank distributions on large graphs.

Next-location recommendation is an emerging task with the proliferation of location-based services. It is the task of recommending the next location to visit for a user, given her past check-in records. Although several principled solutions have been proposed for this task, existing studies have not well characterized the temporal factors in the recommendation. From three real-world datasets, our quantitative analysis reveals that temporal factors play an important role in next-location recommendation, including the periodical temporal preference and dynamic personal preference. In this paper, we propose a Time-Aware Trajectory Embedding Model (TA-TEM) to incorporate three kinds of temporal factors in next-location recommendation. Based on distributed representation learning, the proposed TA-TEM jointly models multiple kinds of temporal factors in a unified manner. TA-TEM also enhances the sequential context by using a longer context window. Experiments show that TA-TEM outperforms several competitive baselines.

Fine-grained classification with few labeled samples has urgent needs in practice since fine-grained samples are more difficult and expensive to collect and annotate. Standard few-shot learning (FSL) focuses on generalising across seen and unseen classes, where the classes are at the same level of granularity. Therefore, when applying existing FSL methods to tackle this problem, large amounts of labeled samples for some fine-grained classes are required. Since samples of coarse-grained classes are much cheaper and easier to obtain, it is desired to learn knowledge from coarse-grained categories that can be transferred to fine-grained classes with a few samples. In this paper, we propose a novel learning problem called cross-granularity few-shot learning (CG-FSL), where sufficient samples of coarse-grained classes are available for training, but in the test stage, the goal is to classify the fine-grained subclasses. This learning paradigm follows the laws of cognitive neurology. We first give an analysis of CG-FSL through the Structural Causal Model (SCM) and figure out that the standard FSL model learned at the coarse-grained level is actually a confounder. We thus perform backdoor adjustment to decouple the interferences and consequently derive a causal CG-FSL model called Meta Attention-Generation Network (MAGN), which is trained in a bilevel optimization manner. We construct benchmarks from several fine-grained image datasets for the CG-FSL problem and empirically show that our model significantly outperforms standard FSL methods and baseline CG-FSL methods.

The emergence of Large Language Models (LLMs) demonstrates their potential to encapsulate the logic and patterns inherent in human behavior simulation by leveraging extensive web data pre-training. However, the boundaries of LLM capabilities in social simulation remain unclear. To further explore the social attributes of LLMs, we introduce the CiteAgent framework, designed to generate citation networks based on human-behavior simulation with LLM-based agents. CiteAgent successfully captures predominant phenomena in real-world citation networks, including power-law distribution, citational distortion, and shrinking diameter. Building on this realistic simulation, we establish two LLM-based research paradigms in social science: LLM-SE (LLM-based Survey Experiment) and LLM-LE (LLM-based Laboratory Experiment). These paradigms facilitate rigorous analyses of citation network phenomena, allowing us to validate and challenge existing theories. Additionally, we extend the research scope of traditional science of science studies through idealized social experiments, with the simulation experiment results providing valuable insights for real-world academic environments. Our work demonstrates the potential of LLMs for advancing science of science research in social science.