No abstract

In this paper, we propose a self-learning architecture for generating natural language templates for conversational assistants. Generating templates to cover all the combinations of slots in an intent is time consuming and labor-intensive. We examine three different models based on our proposed architecture -Rule-based model, Sequence-to-Sequence (Seq2Seq) model and Semantically Conditioned LSTM (SC-LSTM) model for the IoT domain -to reduce the human labor required for template generation. We demonstrate the feasibility of template generation for the IoT domain using our self-learning architecture. In both automatic and human evaluation, the self-learning architecture outperforms previous works trained with a fully human-labeled dataset. This is promising for commercial conversational assistant solutions.

No abstract

Measuring semantic similarity between short texts is challenging because the meaning of short texts may vary dramatically even by a few words due to their limited lengths. In this paper, we propose a novel similarity measure for terms that allows better clustering performance than the state-of-the-art method. To achieve such performance, we incorporate knowledge-based and corpus-based term similarity measures in order to exploit advantages of both approaches. We apply our method to a dialog-utterance dataset, which consists of short dialog texts. Empirical study shows that the proposed method outperforms one of the state-of-the-art clustering algorithms for short text clustering.

No abstract