Search

Recent language generative models are mostly trained on large-scale datasets, while in some real scenarios, the training datasets are often expensive to obtain and would be small-scale. In this paper we investigate the challenging task of less-data constrained generation, especially when the generated news headlines are short yet expected by readers to keep readable and informative simultaneously. We highlight the key information modeling task and propose a novel duality fine-tuning method by formally defining the probabilistic duality constraints between key information prediction and headline generation tasks. The proposed method can capture more information from limited data, build connections between separate tasks, and is suitable for less-data constrained generation tasks. Furthermore, the method can leverage various pre-trained generative regimes, e.g., autoregressive and encoder-decoder models. We conduct extensive experiments to demonstrate that our method is effective and efficient to achieve improved performance in terms of language modeling metric and informativeness correctness metric on two public datasets., Comment: Accepted by AACL-IJCNLP 2022 main conference

Existing Artificial Intelligence (AI) driven automation solutions in enterprises employ machine learning, natural language processing, and chatbots. There is an opportunity for AI Planning to be applied, which offers reasoning about action trajectories to help build automation blueprints. AI Planning is a problem-solving technique, where knowledge about available actions and their consequences is used to identify a sequence of actions, which, when applied in a given initial state, satisfy a desired goal. AI Planning has successfully been applied in a number of domains ranging from space applications, logistics and transportation, manufacturing, robotics, scheduling, e-learning, enterprise risk management, and service composition. In this paper, we discuss experience in building automation solutions that employ AI planning for use in enterprise IT and business services, such as change and event management, migration and transformation and RPA composition. We discuss challenges in adoption of AI planning across the enterprise from implementation and deployment perspectives.

In end-to-end dialogue modeling and agent learning, it is important to (1) effectively learn knowledge from data, and (2) fully utilize heterogeneous information, e.g., dialogue act flow and utterances. However, the majority of existing methods cannot simultaneously satisfy the two conditions. For example, rule definition and data labeling during system design take too much manual work, and sequence-to-sequence methods only model one-side utterance information. In this paper, we propose a novel joint end-to-end model by multi-task representation learning, which can capture the knowledge from heterogeneous information through automatically learning knowledgeable low-dimensional embeddings from data, named with DialogAct2Vec. The model requires little manual work for intervention in system design and we find that the multi-task learning can greatly improve the effectiveness of representation learning. Extensive experiments on a public dataset for restaurant reservation show that the proposed method leads to significant improvements against the state-of-the-art baselines on both the act prediction task and utterance prediction task., Comment: Under review