Task-oriented dialogue systems aim to fulfill user goals through natural language interactions. They are ideally evaluated with human users, which however is unattainable to do at every iteration of the development phase. Simulated users could be an alternative, however their development is nontrivial. Therefore, researchers resort to offline metrics on existing human-human corpora, which are more practical and easily reproducible. They are unfortunately limited in reflecting real performance of dialogue systems. BLEU for instance is poorly correlated with human judgment, and existing corpus-based metrics such as success rate overlook dialogue context mismatches. There is still a need for a reliable metric for task-oriented systems with good generalization and strong correlation with human judgements. In this paper, we propose the use of offline reinforcement learning for dialogue evaluation based on a static corpus. Such an evaluator is typically called a critic and utilized for policy optimization. We go one step further and show that offline RL critics can be trained on a static corpus for any dialogue system as external evaluators, allowing dialogue performance comparisons across various types of systems. This approach has the benefit of being corpus- and model-independent, while attaining strong correlation with human judgements, which we confirm via an interactive user trial.

At the peak of the COVID-19 pandemic, numerous countries worldwide sought to mobilize vaccination campaigns in an attempt to curb the spread and number of deaths caused by the virus. One avenue in which information regarding COVID vaccinations is propagated is that of scientific articles, which provide a certain level of credibility regarding this. Hence, this increases the probability that people who view these articles would get vaccinated if the articles convey a positive message on vaccinations and conversely decreases the probability of vaccinations if the articles convey a negative message. This being said, this study aims to investigate the correlation between article sentiments and the corresponding increase or decrease in vaccinations in the United States. To do this, a lexicon-based sentiment analysis was performed in two steps: first, article content was scraped via a Python library called BeautifulSoup, and second, VADER was used to obtain the sentiment analysis scores for each article based on the scraped text content. Results suggest that there was a relatively weak correlation between the average sentiment score of articles and the corresponding increase or decrease in COVID vaccination rates in the US.

Subjective texts have been especially studied by several works as they can induce certain behaviours in their users. Most work focuses on user-generated texts in social networks, but some other texts also comprise opinions on certain topics and could influence judgement criteria during political decisions. In this work, we address the task of Targeted Sentiment Analysis for the domain of news headlines, published by the main outlets during the 2019 Argentinean Presidential Elections. For this purpose, we present a polarity dataset of 1,976 headlines mentioning candidates in the 2019 elections at the target level. Preliminary experiments with state-of-the-art classification algorithms based on pre-trained linguistic models suggest that target information is helpful for this task. We make our data and pre-trained models publicly available.

Multimodal sentiment analysis is an important research task to predict the sentiment score based on the different modality data from a specific opinion video. Many previous pieces of research have proved the significance of utilizing the shared and unique information across different modalities. However, the high-order combined signals from multimodal data would also help extract satisfied representations. In this paper, we propose CMGA, a Cross-Modality Gated Attention fusion model for MSA that tends to make adequate interaction across different modality pairs. CMGA also adds a forget gate to filter the noisy and redundant signals introduced in the interaction procedure. We experiment on two benchmark datasets in MSA, MOSI, and MOSEI, illustrating the performance of CMGA over several baseline models. We also conduct the ablation study to demonstrate the function of different components inside CMGA.

Zero/few-shot transfer to unseen services is a critical challenge in task-oriented dialogue research. The Schema-Guided Dialogue (SGD) dataset introduced a paradigm for enabling models to support any service in zero-shot through schemas, which describe service APIs to models in natural language. We explore the robustness of dialogue systems to linguistic variations in schemas by designing SGD-X - a benchmark extending SGD with semantically similar yet stylistically diverse variants for every schema. We observe that two top state tracking models fail to generalize well across schema variants, measured by joint goal accuracy and a novel metric for measuring schema sensitivity. Additionally, we present a simple model-agnostic data augmentation method to improve schema robustness.

User simulators (USs) are commonly used to train task-oriented dialogue systems (DSs) via reinforcement learning. The interactions often take place on semantic level for efficiency, but there is still a gap from semantic actions to natural language, which causes a mismatch between training and deployment environment. Incorporating a natural language generation (NLG) module with USs during training can partly deal with this problem. However, since the policy and NLG of USs are optimised separately, these simulated user utterances may not be natural enough in a given context. In this work, we propose a generative transformer-based user simulator (GenTUS). GenTUS consists of an encoder-decoder structure, which means it can optimise both the user policy and natural language generation jointly. GenTUS generates both semantic actions and natural language utterances, preserving interpretability and enhancing language variation. In addition, by representing the inputs and outputs as word sequences and by using a large pre-trained language model we can achieve generalisability in feature representation. We evaluate GenTUS with automatic metrics and human evaluation. Our results show that GenTUS generates more natural language and is able to transfer to an unseen ontology in a zero-shot fashion. In addition, its behaviour can be further shaped with reinforcement learning opening the door to training specialised user simulators.

Recently, some span-based methods have achieved encouraging performances for joint aspect-sentiment analysis, which first extract aspects (aspect extraction) by detecting aspect boundaries and then classify the span-level sentiments (sentiment classification). However, most existing approaches either sequentially extract task-specific features, leading to insufficient feature interactions, or they encode aspect features and sentiment features in a parallel manner, implying that feature representation in each task is largely independent of each other except for input sharing. Both of them ignore the internal correlations between the aspect extraction and sentiment classification. To solve this problem, we novelly propose a hierarchical interactive network (HI-ASA) to model two-way interactions between two tasks appropriately, where the hierarchical interactions involve two steps: shallow-level interaction and deep-level interaction. First, we utilize cross-stitch mechanism to combine the different task-specific features selectively as the input to ensure proper two-way interactions. Second, the mutual information technique is applied to mutually constrain learning between two tasks in the output layer, thus the aspect input and the sentiment input are capable of encoding features of the other task via backpropagation. Extensive experiments on three real-world datasets demonstrate HI-ASA's superiority over baselines.

Multimodal sentiment analysis has become an increasingly popular research area as the demand for multimodal online content is growing. For multimodal sentiment analysis, words can have different meanings depending on the linguistic context and non-verbal information, so it is crucial to understand the meaning of the words accordingly. In addition, the word meanings should be interpreted within the whole utterance context that includes nonverbal information. In this paper, we present a Context-driven Modality Shifting BERT with Contrastive Learning for linguistic, visual, acoustic Representations (CMSBERT-CLR), which incorporates the whole context's non-verbal and verbal information and aligns modalities more effectively through contrastive learning. First, we introduce a Context-driven Modality Shifting (CMS) to incorporate the non-verbal and verbal information within the whole context of the sentence utterance. Then, for improving the alignment of different modalities within a common embedding space, we apply contrastive learning. Furthermore, we use an exponential moving average parameter and label smoothing as optimization strategies, which can make the convergence of the network more stable and increase the flexibility of the alignment. In our experiments, we demonstrate that our approach achieves state-of-the-art results.

Variational Bayes (VB) applied to latent Dirichlet allocation (LDA) has become the most popular algorithm for aspect modeling. While sufficiently successful in text topic extraction from large corpora, VB is less successful in identifying aspects in the presence of limited data. We present a novel variational message passing algorithm as applied to Latent Dirichlet Allocation (LDA) and compare it with the gold standard VB and collapsed Gibbs sampling. In situations where marginalisation leads to non-conjugate messages, we use ideas from sampling to derive approximate update equations. In cases where conjugacy holds, Loopy Belief update (LBU) (also known as Lauritzen-Spiegelhalter) is used. Our algorithm, ALBU (approximate LBU), has strong similarities with Variational Message Passing (VMP) (which is the message passing variant of VB). To compare the performance of the algorithms in the presence of limited data, we use data sets consisting of tweets and news groups. Using coherence measures we show that ALBU learns latent distributions more accurately than does VB, especially for smaller data sets.

Despite having achieved great success for sentiment analysis, existing neural models struggle with implicit sentiment analysis. This may be due to the fact that they may latch onto spurious correlations ("shortcuts", e.g., focusing only on explicit sentiment words), resulting in undermining the effectiveness and robustness of the learned model. In this work, we propose a causal intervention model for Implicit Sentiment Analysis using Instrumental Variable (ISAIV). We first review sentiment analysis from a causal perspective and analyze the confounders existing in this task. Then, we introduce an instrumental variable to eliminate the confounding causal effects, thus extracting the pure causal effect between sentence and sentiment. We compare the proposed ISAIV model with several strong baselines on both the general implicit sentiment analysis and aspect-based implicit sentiment analysis tasks. The results indicate the great advantages of our model and the efficacy of implicit sentiment reasoning.