Radiology report generation systems have the potential to reduce the workload of radiologists by automatically describing the findings in medical images.To broaden the application of the report generation system, the system should generate reports that are not only factually accurate but also chronologically consistent, describing images that are presented in time order, that is, the correct order.We employ a planning-based radiology report generation system that generates the overall structure of reports as "plans" prior to generating reports that are accurate and consistent in order.Additionally, we propose a novel reinforcement learning and inference method, Coordinated Planning (CoPlan), that includes a content planner and a text generator to train and infer in a coordinated manner to alleviate the cascading of errors that are often inherent in planning-based models.We conducted experiments with single-phase diagnostic reports in which the factual accuracy is critical and multi-phase diagnostic reports in which the description order is critical.Our proposed CoPlan improves the content order score by 5.1 pt in time series critical scenarios and the clinical factual accuracy F-score by 9.1 pt in time series irrelevant scenarios, compared those of the baseline models without CoPlan.

Ryuji Kano, Takumi Takahashi, Toru Nishino, Motoki Taniguchi, Tomoki Taniguchi, Tomoko Ohkuma. Proceedings of the 16th Conference of the European Chapter of the Association for Computational Linguistics: Main Volume. 2021.

We present a large-scale corpus of e-mail conversations with domain-agnostic and two-level dialogue act (DA) annotations towards the goal of a better understanding of asynchronous conversations.We annotate over 6,000 messages and 35,000 sentences from more than 2,000 threads.For a domain-independent and application-independent DA annotations, we choose ISO standard 24617-2 as the annotation scheme.To assess the difficulty of DA recognition on our corpus, we evaluate several models, including a pre-trained contextual representation model, as our baselines.The experimental results show that BERT outperforms other neural network models, including previous state-of-the-art models, but falls short of a human performance.We also demonstrate that DA tags of two-level granularity enable a DA recognition model to learn efficiently by using multi-task learning.An evaluation of a model trained on our corpus against other domains of asynchronous conversation reveals the domain independence of our DA annotations.

Increasing users’ positive interactions, such as purchases or clicks, is an important objective of recommender systems. Recommenders typically aim to select items that users will interact with. If the recommended items are purchased, an increase in sales is expected. However, the items could have been purchased even without recommendation. Thus, we want to recommend items that results in purchases caused by recommendation. This can be formulated as a ranking problem in terms of the causal effect. Despite its importance, this problem has not been well explored in the related research. It is challenging because the ground truth of causal effect is unobservable, and estimating the causal effect is prone to the bias arising from currently deployed recommenders. This paper proposes an unbiased learning framework for the causal effect of recommendation. Based on the inverse propensity scoring technique, the proposed framework first constructs unbiased estimators for ranking metrics. Then, it conducts empirical risk minimization on the estimators with propensity capping, which reduces variance under finite training samples. Based on the framework, we develop an unbiased learning method for the causal effect extension of a ranking metric. We theoretically analyze the unbiasedness of the proposed method and empirically demonstrate that the proposed method outperforms other biased learning methods in various settings.

The linear submodular bandit problemwas proposedto simultaneously address diversified retrieval and online learning in a recommender system.If there is no uncertainty, this problem is equivalent toa submodular maximization problem under a cardinality constraint.However, in some situations, recommendation lists should satisfyadditional constraints such as budget constraints,other than a cardinality constraint.Thus, motivated by diversified retrieval considering budget constraints,we introduce a submodular bandit problem under the intersection of$l$ knapsacks and a $k$-system constraint.Here $k$-system constraints forma very general class of constraints includingcardinality constraints and the intersection of $k$ matroid constraints.To solve this problem, we propose a non-greedy algorithm that adaptively focuses ona standard or modified upper-confidence bound.We provide a high-probability upper bound of an approximation regret,where the approximation ratio matches that of a fast offline algorithm.Moreover, we perform experimentsunder various combinations of constraints using asynthetic and two real-world datasetsand demonstrate that our proposed method outperforms the existing baselines.

Toru Nishino, Shotaro Misawa, Ryuji Kano, Tomoki Taniguchi, Yasuhide Miura, Tomoko Ohkuma. Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing and the 9th International Joint Conference on Natural Language Processing (EMNLP-IJCNLP). 2019.

Recommender systems aim to increase user actions such as clicks and purchases. Typical evaluations of recommenders regard the purchase of a recommended item as a success. However, the item may have been purchased even without the recommendation. An uplift is defined as an increase in user actions caused by recommendations. Situations with and without a recommendation cannot both be observed for a specific user-item pair at a given time instance, making uplift-based evaluation and optimization challenging. This paper proposes new evaluation metrics and optimization methods for the uplift in a recommender system. We apply a causal inference framework to estimate the average uplift for the offline evaluation of recommenders. Our evaluation protocol leverages both purchase and recommendation logs under a currently deployed recommender system, to simulate the cases both with and without recommendations. This enables the offline evaluation of the uplift for newly generated recommendation lists. For optimization, we need to define positive and negative samples that are specific to an uplift-based approach. For this purpose, we deduce four classes of items by observing purchase and recommendation logs. We derive the relative priorities among these four classes in terms of the uplift and use them to construct both pointwise and pairwise sampling methods for uplift optimization. Through dedicated experiments with three public datasets, we demonstrate the effectiveness of our optimization methods in improving the uplift.

Computer-aided education systems are now seeking to provide each student with personalized materials based on a student's individual knowledge. To provide suitable learning materials, tracing each student's knowledge over a period of time is important. However, predicting each student's knowledge is difficult because students tend to forget. The forgetting behavior is mainly because of two reasons: the lag time from the previous interaction, and the number of past trials on a question. Although there are a few studies that consider forgetting while modeling a student's knowledge, some models consider only partial information about forgetting, whereas others consider multiple features about forgetting, ignoring a student's learning sequence. In this paper, we focus on modeling and predicting a student's knowledge by considering their forgetting behavior. We extend the deep knowledge tracing model [17], which is a state-of-the-art sequential model for knowledge tracing, to consider forgetting by incorporating multiple types of information related to forgetting. Experiments on knowledge tracing datasets show that our proposed model improves the predictive performance as compared to baselines. Moreover, we also examine that the combination of multiple types of information that affect the behavior of forgetting results in performance improvement.

We proposed a model that integrates discussion structures with neural networks to classify discourse acts. Several attempts have been made in earlier works to analyze texts that are used in various discussions. The importance of discussion structures has been explored in those works but their methods required a sophisticated design to combine structural features with a classifier. Our model introduces tree learning approaches and a graph learning approach to directly capture discussion structures without structural features. In an evaluation to classify discussion discourse acts in Reddit, the model achieved improvements of 1.5% in accuracy and 2.2 in FB1 score compared to the previous best model. We further analyzed the model using an attention mechanism to inspect interactions among different learning approaches.

We leverage a popularity measure in social media as a distant label for extractive summarization of online conversations.In social media, users can vote, share, or bookmark a post they prefer.The number of these actions is regarded as a measure of popularity.However, popularity is not determined solely by content of a post, e.g., a text or an image it contains, but is highly based on its contexts, e.g., timing, and authority.We propose Disjunctive model that computes the contribution of content and context separately.For evaluation, we build a dataset where the informativeness of comments is annotated.We evaluate the results with ranking metrics, and show that our model outperforms the baseline models which directly use popularity as a measure of informativeness.

Recommender systems support user decision-making, and explanations of recommendations further facilitate their usefulness. Previous explanation styles are based on similar users, similar items, demographics of users, and contents of items. Contexts, such as usage scenarios and accompanying persons, have not been used for explanations, although they influence user decisions. In this paper, we propose a context style explanation method, presenting contexts suitable for consuming recommended items. The expected impacts of context style explanations are 1) persuasiveness: recognition of suitable context for usage motivates users to consume items, and 2) usefulness: envisioning context helps users to make right choices because the values of items depend on contexts. We evaluate context style persuasiveness and usefulness by a crowdsourcing-based user study in a restaurant recommendation setting. The context style explanation is compared to demographic and content style explanations. We also combine context style and other explanation styles, confirming that hybrid styles improve persuasiveness and usefulness of explanation.

We propose a novel geolocation prediction model using a complex neural network.Our model unifies text, metadata, and user network representations with an attention mechanism to overcome previous ensemble approaches.In an evaluation using two open datasets, the proposed model exhibited a maximum 3.8% increase in accuracy and a maximum of 6.6% increase in accuracy@161 against previous models.We further analyzed several intermediate layers of our model, which revealed that their states capture some statistical characteristics of the datasets.