Incorrect information poses significant challenges by disrupting content veracity and integrity, yet most detection approaches struggle to jointly balance textual content verification with external knowledge modification under collapsed attention geometries.To address this issue, we propose a dual-head reasoning framework, BiMind, which disentangles content-internal reasoning from knowledgeaugmented reasoning.In BiMind, we introduce three core innovations: (i) an attention geometry adapter that reshapes attention logits via token-conditioned offsets and mitigates attention collapse; (ii) a self-retrieval knowledge mechanism, which constructs an in-domain semantic memory through kNN retrieval and injects retrieved neighbors via feature-wise linear modulation; (iii) the uncertainty-aware fusion strategies, including entropy-gated fusion and a trainable agreement head, stabilized by a symmetric Kullback-Leibler agreement regularizer.To quantify the knowledge contributions, we define a novel metric, Value-of-eXperience (VoX), to measure instance-wise logit gains from knowledge-augmented reasoning.Experiment results on public datasets demonstrate that our BiMind model outperforms advanced detection approaches and provides interpretable diagnostics on when and why knowledge matters.Our BiMind model and the tested datasets are available at https://github.com/cvzh/BiMind.

Social media advertising has become one of the most popular forms of advertising in recent years. Advancements in video and image generation have led to the creation of avatars, which have changed the landscape of online marketing. The avatars used for social media advertising are known as virtual influencers. In this paper, we use Instagram comments to observe consumer perceptions of virtual influencers (VIs) and human influencers (HIs) in social media advertising. By investigating how consumers engage with and respond to VIs, the study aims to provide insights into their effectiveness compared to HIs. In this study, we consider two timelines, pre- and post-pandemic, for different types of posts by VIs and HIs. We focus mainly on two user-affective states: sentiment and skepticism. The results show that VIs are perceived with more negativity and skepticism compared to HIs, irrespective of the type of post in both time periods. However, negative and skeptical responses towards VIs decreased post-pandemic. This empirical study provides insights into the evolving field of influencer advertising and highlights changes in consumer perception pre- and post-pandemic towards VIs and HIs.

Individuals often encounter persuasion attempts, during which a persuasion agent aims to persuade a target to change the target’s emotions, beliefs, and behaviors. These persuasion attempts can be observed in various social settings, such as advertising, public health, political campaigns, and personal relationships. During these persuasion attempts, targets generally like to preserve their autonomy, so their responses often manifest in some form of resistance, like a skeptical reaction. In order to detect such skepticism in response to persuasion attempts on social media, we developed a corpus based on consumer psychology. In this paper, we consider one of the most prominent areas in which persuasion attempts unfold: social media influencer marketing. In this paper, we introduce the skepticism detection corpus, SkOTaPA, which was developed using multiple independent human annotations, and inter-coder reliability was evaluated with Krippendorff’s alpha (0.709). We performed validity tests to show skepticism cannot be detected using other potential proxy variables like sentiment and sarcasm.

The increase in analysis of real life social networks has led to a better understanding of the ways humans socialize in a group. Since trust is an important part of any social interaction, researchers use such networks to understand the nuances of trust relationships. One of the major requirements in trust applications is identifying the trustworthy actors in these networks. This paper proposes a pair of complementary measures that can be used to measure trust scores of actors in a social network using involvement of social networks. Based on the proposed measures, an iterative matrix convergence algorithm is developed that calculates the trustingness and the trustworthiness of each actor in the network. Trustingness of an actor is defined as the propensity of an actor to trust his neighbors in the network. Trustworthiness, on the other hand, is defined as the willingness of the network to trust an individual actor. The algorithm is proposed based on the idea that a person having higher trustingness score contributes to the trustworthiness of its neighbors to a lower degree. Conversely, a higher trustworthiness score is a result of lots of neighbors linked to the actor having low trustingness scores. The algorithm runs in O(k × |E|) time where k denotes the number of iterations and |E| denotes the number of edges in the network. Moreover, the paper shows that the algorithm converges to a finite value quickly. Finally the proposed scores for trust prediction is implemented for various social networks and is shown that the proposed algorithm performs better (average 5%) than the state of the art trust scoring algorithms. The full version of the paper along with the coding implementation can be found at [3].