Class incremental learning (CIL) devotes to addressing catastrophic forgetting while continually learning new tasks. Recently, prompt tuning techniques based on vision transformers (ViT) have achieved promising results in rehearsal-free CIL. To alleviate forgetting, representative methods use a query-key mechanism to generate prompts and attach them to the frozen pre-trained ViT. However, these methods neglect the effect of query, and the learning capacity of the model is limited due to unsuitable prompts. In this paper, we propose a new approach called Prompting to Prompt (P2P). Instead of using a task-independent query function, we learn sample queries together with prompts in response to the shift of data distribution in CIL. P2P can better separate classes across tasks because the generated prompts are effective and more discriminative sample features can be extracted. Besides, the whole training process is end-to-end and queries are decided by prompts themselves, which avoids additional parameters. P2P improves the plasticity of model while maintaining good resistance to forgetting in the long task sequence. Experiments show that our approach achieves state-of-the-art results with even fewer parameters.

Cross-domain few-shot classification problem is one of the most challenging problems in few-shot learning, as it requires transferring useful information from the source dataset, which has large distribution differences, to downstream few-shot tasks. Most previous methods focused on improving model representation from the perspective of model structure while ignoring the impact of data distribution. Our exploratory experiment indicates that different categories of data in the source dataset have different impacts on downstream tasks. This signifies that we should adaptively adjust the contribution of different categories of training data based on downstream tasks. Inspired by this, we design a novel training framework, which allows us to disentangle the information of different sample clusters from a well-trained network and adaptively aggregate the information to assist the downstream tasks. Our comprehensive experiments on four commonly used datasets verify that our method can improve classification accuracy while converging with fewer training epochs during fine-tuning.

Class incremental learning devotes to learning a classification model from incrementally arriving training data. Existing methods tend to use a single-headed layout due to the lack of task delimiter while testing. However, this is not suitable for Incremental Implicitly-Refined Classification (IIRC), an extension to class incremental learning where different classes could have two granularity levels. In IIRC, each sample could have a coarse label and a fine label. Without considering hierarchical relations among classes, it's difficult to distinguish a subclass from its siblings and assign them to the same superclass while alleviating catastrophic forgetting. In this paper, we propose a new framework called Mutually Promoted Hierarchical Learning (MPHL) to solve IIRC. MPHL learns separate latent spaces for superclasses and subclasses to coordinate the representations of two granularities and utilizes the hierarchy to facilitate the training process of each other. While using knowledge distillation, subclasses are treated as replay samples of their parents, and coarse labels are used to eliminate interference from non-siblings. To reduce the risk of selecting unwarranted fine-grained labels, we further propose a self-adaptive threshold strategy to detect out-of-distribution samples. Extensive experiments on IIRC-CIFAR100 and IIRC-ImageNet show that MPHL achieves state-of-the-art results.

We consider the problem of class incremental learning (CIL), where an agent aims to learn new classes continually without forgetting previous ones. As one of the mainstream paradigms of incremental learning, parameter isolation methods prevent forgetting by allocating different model parameters to each task, but knowledge transfer across tasks is difficult and usually overlooked. As a consequence, the discriminability between old and new classes is limited, especially when training data of old classes is not accessible. In this paper, we propose a new data-free approach named Twin Contrastive Networks (TCN) for CIL by utilizing the connections among tasks and network parameters. Specifically, we treat CIL as a sequence of one-class classification tasks and train separate classifiers to identify each class. To facilitate knowledge transfer and make full use of accumulated knowledge, a twin network structure is adopted to learn different feature representations for future use. While encountering new classes, previous twin networks are utilized directly by a contrastive loss to improve the model's discriminability. TCN avoids catastrophic forgetting by fixing all learnt parameters and leverages prior knowledge contained in networks. Experiments on three widely used incremental learning benchmarks verify the effectiveness of TCN.

Social recommendation provides a promising way to alleviate data sparsity and boost recommendation performance in collaborative filtering. How to model social influence is a central problem in the social recommendation. The social data can be represented by two separate graphs together, i.e., a social (user-user) graph and an interaction (user-item) graph. To some extent, this representation leads to loss of heterogeneity. To address this problem, we propose a novel social recommendation model called Heterogeneous Information Diffusion Model (HIDM for short). Specifically, since users are involved in both two subspaces: social subspace and interaction subspace, we design a neural architecture with two modules to aggregate information of the two subspaces. The first module is social aggregation which models the social influence from social connections in layer-wise via GNNs. The second module is item aggregation which captures the item influence from interacted items in a single layer. Moreover, different message functions are designed to encode information in different subspaces. Finally, the user preference representation is obtained by fusing the influence of the two subspaces. The extensive experiments conducted on three realworld datasets show that HIDM outperforms several different state-of-the-art social recommendation models.

Heterogeneous information network (HIN) based models aim to make recommendations as much as possible using meta-path information. It would be more advisable if they consider the scenario information provided by meta-path and explore the relatedness of meta-paths. To address these issues, we propose an End-to-end model based on Meta-Embedding for the Recommendation over HIN, called EMER. Here, meta-embedding refers to the explicit representation of meta-path under the user and item pair. Learning the meta-embedding can capture the scenario information provided by meta-path, and can directly explore the relatedness between meta-paths. Furthermore, we leverage the meta-embedding to design three loss functions of structure loss, relation loss and rating loss to capture three aspect information. We use structure loss to calculate the similarity between nodes based on meta-embedding to approximate the similarity calculated based on meta-path for preserving the structural characteristics. The relation loss models meta-paths by interpreting them as translations operating on the embeddings of users and items. We use rating loss to characterize the difference between the predicted ratings and the truth. Experiments over three real datasets indicate that our model achieves state-of-the-art performance.

As one of fundamental properties to characterize inconsistency measures for knowledge bases, the property of free formula independence captures well the intuition that free formulas are independent of the amount of inconsistency in a knowledge base for cases where inconsistency is characterized in terms of minimal inconsistent subsets. But it has been argued that not all the free formulas are independent of inconsistency in some other contexts of inconsistency characterization. In this paper, we propose a notion of Bi-free formula to describe formulas that are free from inconsistency in both syntactic characterization and paraconsistent models in the framework of Priest's minimally inconsistent LP. Then we propose the property of Bi-free formula independence, which is more suitable for characterizing the role of formulas free from inconsistency in measuring inconsistency from both syntactic and semantic perspectives.

Most matrix factorization methods for rating prediction in recommender system seek to model the preference of each user by a latent feature vector with small dimension. Moreover, these methods assume that user's preferences over all items may be captured by the same vector. However, a given user may have different preferences on a given item when he/she plays different roles in different scenarios. In this paper, we propose a new matrix factorization model named PSVD, which allows us to capture user's different preferences over different items flexibly in rating prediction. Specially, we use a pair of preferences to represent the whole preference of user over items. Then the dual preferences are considered simultaneously in building the latent feature vector of user. Moreover, PSVD model allows users to adjust their own feature vector when selecting different products. A comprehensive set of experiments on three public datasets show that PSVD model can outperform several models in terms of different evaluation metrics.

Network embedding, which learns continuous low-dimension representations of nodes, provides an effective way for many network analysis tasks, such as node classification and link prediction. Most existing models are time-consuming and cannot be applied to dynamic networks. To alleviate these issues, we apply Arora's sentence2vec model to network embedding to enhance the performance of existing network embedding methods. Under the same framework of the sentence2vec, we name the network embedding method MNE^2, which allows us to leverage the latent representation obtained from a given embedding approach to learn enhanced node embeddings in a network. Taking into account interactions between nodes in the network, the enhanced network embedding can be viewed as the latent factors for generating embedding representations of neighbor nodes. Then, PCA is applied to modify the embedding results to make the enhanced embeddings more expressive. We evaluate MNE^2 on three real-world social network datasets for node classification and link prediction tasks. The results show that MNE^2 can outperform state-of-the-art network embedding learning methods in both tasks.

As one of fundamental properties to characterize inconsistency measures for knowledge bases, the property of free formula independence well captures the intuition that free formulas are independent of the amount of inconsistency in a knowledge base for cases where inconsistency is characterized in terms of minimal inconsistent subsets. But it has been argued that not all the free formulas are independent of inconsistency in some other contexts of inconsistency characterization. In this paper, we propose a characterization of formulas independent of inconsistency in the framework of Priest's minimally inconsistent LP. Based on an atom-based counterpart of the notion of free formula, we propose a notion of Bi-free formula to describe formulas that are free from inconsistency in both syntax and paraconsistent models in this logic. Then we propose the property of Bi-free formula independence, which is more suitable for characterizing the role of formulas free from inconsistency in measuring inconsistency from both syntactic and semantic perspectives.

Abstract This paper proposes a model, the linear model, for randomly generating logic programs with low density of rules and investigates statistical properties of such random logic programs. It is mathematically shown that the average number of answer sets for a random program converges to a constant when the number of atoms approaches infinity. Several experimental results are also reported, which justify the suitability of the linear model. It is also experimentally shown that, under this model, the size distribution of answer sets for random programs tends to a normal distribution when the number of atoms is sufficiently large.

It is increasingly recognized that identifying the degree of blame or responsibility of each formula for inconsistency of a knowledge base (i.e. a set of formulas) is useful for making rational decisions to resolve inconsistency in that knowledge base. Most current techniques for measuring the blame of each formula with regard to an inconsistent knowledge base focus on classical knowledge bases only. Proposals for measuring the blames of formulas with regard to an inconsistent prioritized knowledge base have not yet been given much consideration. However, the notion of priority is important in inconsistency-tolerant reasoning. This article investigates this issue and presents a family of measurements for the degree of blame of each formula in an inconsistent prioritized knowledge base by using the minimal inconsistent subsets of that knowledge base. First of all, we present a set of intuitive postulates as general criteria to characterize rational measurements for the blames of formulas of an inconsistent prioritized knowledge base. Then we present a family of measurements for the blame of each formula in an inconsistent prioritized knowledge base under the guidance of the principle of proportionality, one of the intuitive postulates. We also demonstrate that each of these measurements possesses the properties that it ought to have. Finally, we use a simple but explanatory example in requirements engineering to illustrate the application of these measurements. Compared to the related works, the postulates presented in this article consider the special characteristics of minimal inconsistent subsets as well as the priority levels of formulas. This makes them more appropriate to characterizing the inconsistency measures defined from minimal inconsistent subsets for prioritized knowledge bases as well as classical knowledge bases. Correspondingly, the measures guided by these postulates can intuitively capture the inconsistency for prioritized knowledge bases.

Developing a desirable framework for handling inconsistencies in software requirements specifications is a challenging problem. It has been widely recognized that the relative priority of requirements can help developers to make some necessary trade-off decisions for resolving con- flicts. However, for most distributed development such as viewpoints-based approaches, different stakeholders may assign different levels of priority to the same shared requirements statement from their own perspectives. The disagreement in the local levels of priority assigned to the same shared requirements statement often puts developers into a dilemma during the inconsistency handling process. The main contribution of this paper is to present a prioritized merging-based framework for handling inconsistency in distributed software requirements specifications. Given a set of distributed inconsistent requirements collections with the local prioritization, we first construct a requirements specification with a prioritization from an overall perspective. We provide two approaches to constructing a requirements specification with the global prioritization, including a merging-based construction and a priority vector-based construction. Following this, we derive proposals for handling inconsistencies from the globally prioritized requirements specification in terms of prioritized merging. Moreover, from the overall perspective, these proposals may be viewed as the most appropriate to modifying the given inconsistent requirements specification in the sense of the ordering relation over all the consistent subsets of the requirements specification. Finally, we consider applying negotiation-based techniques to viewpoints so as to identify an acceptable common proposal from these proposals.

Inconsistency is a pervasive issue in software engineering. Both general rules of inconsistency management and special case-based approaches to handling inconsistency have recently been considered. In this paper, we present a process model for handling requirements inconsistency within the viewpoints framework. In this process model, when an inconsistency among viewpoints is detected, a set of candidate proposals for handling inconsistency will be generated using techniques from multi-agent automated negotiations. The proposals are then prioritized using an integrated measurement of inconsistencies. The viewpoints involved in the inconsistency will then enter the negotiations by being presented with the candidate proposals and thus selecting an acceptable proposal based on the priorities associated with each candidate proposal. To facilitate usability, in our process, we assume that the natural language requirements statements are first translated into corresponding logical formulas using a translator software. Moreover, the candidate proposals for handling inconsistency are also translated back from formal logic into natural language before being presented for selection.