Detecting Out-of-Domain (OOD) or unknown intents from user queries is essential in a taskoriented dialog system. A key challenge of OOD detection is to learn discriminative semantic features. Traditional cross-entropy loss only focuses on whether a sample is correctly classified, and does not explicitly distinguish the margins between categories. In this paper, we propose a supervised contrastive learning objective to minimize intra-class variance by pulling together in-domain intents belonging to the same class and maximize inter-class variance by pushing apart samples from different classes. Besides, we employ an adversarial augmentation mechanism to obtain pseudo diverse views of a sample in the latent space. Experiments on two public datasets prove the effectiveness of our method capturing discriminative representations for OOD detection. 1
Purpose The purpose of this paper is to study the participation behaviors in the context of crowdsourcing projects from the perspective of gamification. Design/methodology/approach This paper first proposed a model to depict the effect of four categories of game elements on three types of motivation based upon several motivation theories, which may, in turn, influence user participation. Then, 5 × 2 between-subject Web experiments were designed for collecting data and validating this model. Findings Game elements which provide participants with rewards and recognitions or remind participants of the completion progress of their tasks may positively influence the extrinsic motivation, whereas game elements which can help create a fantasy scene may strengthen intrinsic motivation. Besides, recognition-kind and progress-kind game elements may trigger the internalization of extrinsic motivation. In addition, when a task is of high complexity, the effects from game elements on extrinsic motivation and intrinsic motivation will be less prominent, whereas the internalization of extrinsic motivation may benefit from the increase of task complexity. Originality/value This study may uncover the motivation mechanism of several different kinds of game elements, which may help to find which game elements are more effective in enhancing engagement and participation in crowdsourcing projects. Besides, as task complexity is used as a moderator, one may be able to identify whether task complexity is able to influence the effects from game elements on motivations. Last, but not the least, this study will indicate the interrelationship between game elements, individual motivation and user participation, which can be adapted by other scholars.
Detecting out-of-domain (OOD) intents is crucial for the deployed task-oriented dialogue system. Previous unsupervised OOD detection methods only extract discriminative features of different in-domain intents while supervised counterparts can directly distinguish OOD and in-domain intents but require extensive labeled OOD data. To combine the benefits of both types, we propose a selfsupervised contrastive learning framework to model discriminative semantic features of both in-domain intents and OOD intents from unlabeled data. Besides, we introduce an adversarial augmentation neural module to improve the efficiency and robustness of contrastive learning. Experiments on two public benchmark datasets show that our method can consistently outperform the baselines with a statistically significant margin.
Consumers’ purchase behavior increasingly relies on online reviews. Accordingly, there are more and more deceptive reviews which are harmful to customers. Existing methods to detect spam reviews mainly take the problem as a general text classification task, but they ignore the important features of spam reviews. In this paper, we propose a novel model, which splits a review into three parts: first sentence, middle context, and last sentence, based on the discovery that the first and last sentence express stronger emotion than the middle context. Then, the model uses four independent bidirectional long-short term memory (LSTM) models to encode the beginning, middle, end of a review and the whole review into four document representations. After that, the four representations are integrated into one document representation by a self-attention mechanism layer and an attention mechanism layer. Based on three domain datasets, the results of in-domain and mix-domain experiments show that our proposed method performs better than the compared methods.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.