Zero-shot prompt-based learning has made much progress in sentiment analysis, and considerable effort has been dedicated to designing high-performing prompt templates. However, two problems exist; First, large language models are often biased to their pre-training data, leading to poor performance in prompt templates that models have rarely seen. Second, in order to adapt to different domains, re-designing prompt templates is usually required, which is time-consuming and inefficient. To remedy both shortcomings, we propose a simple yet strong data construction method to de-bias a given prompt template, yielding a large performance improvement in sentiment analysis tasks across different domains, pre-trained language models, and prompt templates. Also, we demonstrate the advantage of using domain-agnostic generic responses over the in-domain ground-truth data.

Syntactic knowledge is invaluable information for many tasks which handle complex or long sentences, but typical pre-trained language models do not contain sufficient syntactic knowledge. Thus it results in failures in downstream tasks that require syntactic knowledge. In this paper, we explore additional training to incorporate syntactic knowledge to a language model. We designed four pre-training tasks that learn different syntactic perspectives. For adding new syntactic knowledge and keeping a good balance between the original and additional knowledge, we addressed the problem of catastrophic forgetting that prevents the model from keeping semantic information when the model learns additional syntactic knowledge. We demonstrated that additional syntactic training produced consistent performance gains while clearly avoiding catastrophic forgetting.

Deletion-based sentence compression in the English language has made significant progress over the past few decades. However, there is a lack of large-scale and high-quality parallel corpus (i.e., (sentence, compression) pairs) for the Chinese language to train an efficient compression system. To remedy this shortcoming, we present a dependency-tree-based method to construct a Chinese corpus with 151k pairs of sentences and compression based on Chinese language-specific characteristics. Subsequently, we trained both extractive and generative neural compression models using the constructed corpus. The experimental results show that our compression model can generate high-quality compressed sentences on both automatic and human evaluation metrics compared with the baselines. The results of the faithfulness evaluation also indicated that the Chinese compression model trained on our constructed corpus can produce more faithful compressed sentences. Furthermore, a dataset with 1,000 pairs of sentences and ground truth compression was manually created for automatic evaluation, which, we believe, will benefit future research on Chinese sentence compression.

We propose a new word representation method derived from visual objects in associated images to tackle the lexical entailment task. Although it has been shown that the Distributional Informativeness Hypothesis (DIH) holds on text, in which the DIH assumes that a context surrounding a hyponym is more informative than that of a hypernym, it has never been tested on visual objects. Since our perception is tightly associated with language, it is meaningful to explore whether the DIH holds on visual objects. To this end, we consider visual objects as the context of a word and represent a word as a bag of visual objects found in images associated with the word. This allows us to test the feasibility of the visual DIH. To better distinguish word pairs in a hypernym relation from other relations such as co-hypernyms, we also propose a new measurable function that takes into account both the difference in the generality of meaning and similarity of meaning between words. Our experimental results show that the DIH holds on visual objects and that the proposed method combined with the proposed function outperforms existing unsupervised representation methods.

Conventional multimodal tasks, such as caption generation and visual question answering, have allowed machines to understand an image by describing or being asked about it in natural language, often via a sentence. Datasets for these tasks contain a large number of pairs of an image and the corresponding sentence as an instance. However, a real multimodal document such as a news article or Wikipedia page consists of multiple sentences with multiple images. Such documents require an advanced skill of jointly considering the multiple texts and multiple images, beyond a single sentence and image, for the interpretation. Therefore, aiming at building a system that can understand multimodal documents, we propose a task called image position prediction (IPP). In this task, a system learns plausible positions of images in a given document. To study this task, we automatically constructed a dataset of 66K multimodal documents with 320K images from Wikipedia articles. We conducted a preliminary experiment to evaluate the performance of a current multimodal system on our task. The experimental results show that the system outperformed simple baselines while the performance is still far from human performance, which thus poses new challenges in multimodal research.

This paper demonstrates a neural parser implementation suitable for consistently head-final languages such as Japanese. Unlike the transition- and graph-based algorithms in most state-of-the-art parsers, our parser directly selects the head word of a dependent from a limited number of candidates. This method drastically simplifies the model so that we can easily interpret the output of the neural model. Moreover, by exploiting grammatical knowledge to restrict possible modification types, we can control the output of the parser to reduce specific errors without adding annotated corpora. The neural parser performed well both on conventional Japanese corpora and the Japanese version of Universal Dependency corpus, and the advantages of distributed representations were observed in the comparison with the non-neural conventional model.

This paper presents our system submitted for the CoNLL 2017 Shared Task, “Multilingual Parsing from Raw Text to Universal Dependencies.” We ran the system for all languages with our own fully pipelined components without relying on re-trained baseline systems. To train the dependency parser, we used only the universal part-of-speech tags and distance between words, and applied deterministic rules to assign dependency labels. The simple and delexicalized models are suitable for cross-lingual transfer approaches and a universal language model. Experimental results show that our model performed well in some metrics and leads discussion on topics such as contribution of each component and on syntactic similarities among languages.