Intent classification is an important building block of dialogue systems. With the burgeoning of conversational AI, existing systems are not capable of handling numerous fast-emerging intents, which motivates zeroshot intent classification. Nevertheless, research on this problem is still in the incipient stage and few methods are available. A recently proposed zero-shot intent classification method, IntentCapsNet, has been shown to achieve state-of-the-art performance. However, it has two unaddressed limitations: (1) it cannot deal with polysemy when extracting semantic capsules; (2) it hardly recognizes the utterances of unseen intents in the generalized zero-shot intent classification setting. To overcome these limitations, we propose to reconstruct capsule networks for zeroshot intent classification. First, we introduce a dimensional attention mechanism to fight against polysemy. Second, we reconstruct the transformation matrices for unseen intents by utilizing abundant latent information of the labeled utterances, which significantly improves the model generalization ability. Experimental results on two task-oriented dialogue datasets in different languages show that our proposed method outperforms IntentCapsNet and other strong baselines.

No abstract

An increasing number of social media users are becoming used to disseminate activities through geotagged posts. The massive available geotagged posts enable collections of users' footprints over time and offer effective opportunities for mobility prediction. Using geotagged posts for spatio-temporal prediction of future location, however, is challenging. Previous studies either focus on nextplace prediction or rely on dense data sources such as GPS data. Introduced in this article is a novel method for future location prediction of individuals based on geotagged social media data. This method employs the hierarchical densitybased clustering algorithm with adaptive parameter selection to identify the regions frequently visited by a social media user. A multi-feature weighted Bayesian model is then developed to forecast users' spatio-temporal locations by combining multiple factors affecting human mobility patterns. Further, an updating strategy is designed to efficiently adjust, over time, the proposed model to the dynamics in users' mobility patterns. Based on two real-life datasets, the proposed approach outperforms a state-of-the-art method in prediction accuracy by up to 5.34% and 3.30%. Tests show prediction reliability is high with quality predictions, but low in the identification of erroneous locations.

No abstract

End-to-end Spoken Language Understanding (E2E SLU) has attracted increasing interest due to its advantages of joint optimization and low latency when compared to traditionally cascaded pipelines. Existing E2E SLU models usually follow a two-stage configuration where an Automatic Speech Recognition (ASR) network first predicts a transcript which is then passed to a Natural Language Understanding (NLU) module through an interface to infer semantic labels, such as intent and slot tags. This design, however, does not consider the NLU posterior while making transcript predictions, nor correct the NLU prediction error immediately by considering the previously predicted word-pieces. In addition, the NLU model in the two-stage system is not streamable, as it must wait for the audio segments to complete processing, which ultimately impacts the latency of the SLU system. In this work, we propose a streamable multi-task semantic transducer model to address these considerations. Our proposed architecture predicts ASR and NLU labels auto-regressively and uses a semantic decoder to ingest both previously predicted word-pieces and slot tags while aggregating them through a fusion network. Using an industry scale SLU and a public FSC dataset, we show the proposed model outperforms the two-stage E2E SLU model for both ASR and NLU metrics.