In this study, we propose a morpheme-based scheme for Korean dependency parsing and adopt the proposed scheme to Universal Dependencies. We present the linguistic rationale that illustrates the motivation and the necessity of adopting the morpheme-based format, and develop scripts that convert between the original format used by Universal Dependencies and the proposed morpheme-based format automatically. The effectiveness of the proposed format for Korean dependency parsing is then testified by both statistical and neural models, including UDPipe and Stanza, with our carefully constructed morpheme-based word embedding for Korean. morphUD outperforms parsing results for all Korean UD treebanks, and we also present detailed error analyses.

Multilingual pre-trained language models have shown impressive performance on cross-lingual tasks. It greatly facilitates the applications of natural language processing on low-resource languages. However, there are still some languages that the current multilingual models do not perform well on. In this paper, we propose CINO (Chinese Minority Pre-trained Language Model), a multilingual pre-trained language model for Chinese minority languages. It covers Standard Chinese, Yue Chinese, and six other ethnic minority languages. To evaluate the cross-lingual ability of the multilingual model on ethnic minority languages, we collect documents from Wikipedia and news websites, and construct two text classification datasets, WCM (Wiki-Chinese-Minority) and CMNews (Chinese-Minority-News). We show that CINO notably outperforms the baselines on various classification tasks. The CINO model and the datasets are publicly available at http://cino.hfl-rc.com.

Entity alignment (EA) aims to find entities in different knowledge graphs (KGs) that refer to the same object in the real world. Recent studies incorporate temporal information to augment the representations of KGs. The existing methods for EA between temporal KGs (TKGs) utilize a time-aware attention mechanism to incorporate relational and temporal information into entity embeddings. The approaches outperform the previous methods by using temporal information. However, we believe that it is not necessary to learn the embeddings of temporal information in KGs since most TKGs have uniform temporal representations. Therefore, we propose a simple graph neural network (GNN) model combined with a temporal information matching mechanism, which achieves better performance with less time and fewer parameters. Furthermore, since alignment seeds are difficult to label in real-world applications, we also propose a method to generate unsupervised alignment seeds via the temporal information of TKG. Extensive experiments on public datasets indicate that our supervised method significantly outperforms the previous methods and the unsupervised one has competitive performance.

Multi-agent systems for resource allocation (MRAs) have been introduced as a concept for modelling competitive resource allocation problems in distributed computing. An MRA is composed of a set of agents and a set of resources. Each agent has goals in terms of allocating certain resources. For MRAs it is typically of importance that they are designed in a way such that there exists a strategy that guarantees that all agents will achieve their goals. The corresponding model checking problem is to determine whether such a winning strategy exists or not, and the synthesis problem is to actually build the strategy. While winning strategies ensure that all goals will be achieved, following such strategies does not necessarily involve an optimal use of resources. In this paper, we present a technique that allows to synthesise cost-optimal solutions to distributed resource allocation problems. We consider a scenario where system components such as agents and resources involve costs. A multi-agent system shall be designed that is cost-minimal but still capable of accomplishing a given set of goals. Our approach synthesises a winning strategy that minimises the cumulative costs of the components that are required for achieving the goals. The technique is based on a propositional logic encoding and a reduction of the synthesis problem to the maximum satisfiability problem (Max-SAT). Hence, a Max-SAT solver can be used to perform the synthesis. From a truth assignment that maximises the number of satisfied clauses of the encoding a cost-optimal winning strategy as well as a cost-optimal system can be immediately derived.

This paper measures the stability of cross-linguistic register variation. A register is a variety of a language that is associated with extra-linguistic context. The relationship between a register and its context is functional: the linguistic features that make up a register are motivated by the needs and constraints of the communicative situation. This view hypothesizes that register should be universal, so that we expect a stable relationship between the extra-linguistic context that defines a register and the sets of linguistic features which the register contains. In this paper, the universality and robustness of register variation is tested by comparing variation within vs. between registerspecific corpora in 60 languages using corpora produced in comparable communicative situations: tweets and Wikipedia articles. Our findings confirm the prediction that register variation is, in fact, universal. A variety of a language is a combination of linguistic features that co-vary together: for example, past tenses and third person pronouns, nouns and determiners. A register can be defined as a variety of a language that is associated with a specific context of production (Biber and Conrad 2009). In this way, registers contrast with other types of varieties, such as dialects or sociolects, which are instead associated with social factors. The relationship between a register and its context is functional in nature: for example, the features of a particular register are used because they respond to the constraints and needs of that situation. For example, the past tense and third person pronouns are tools we need to construct a narrative and their usage therefore correlates with situations in which one of the purposes is to narrate (e.g. a fictional novel but also a biography). In the same way, nominalisations and the passive voice can be useful to remove agency from a text, thus being quite useful in scientific and academic prose. This deep connection between a register and its context means that both need to be described in order to carry out a register analysis. The language of the register is described by referring to linguistic features, which tend to be lexicogrammatical items. And the context of production tends to be described through an analysis of its contextual configuration, for example using Situational Parameters (Biber 1994; Biber and Conrad 2009), a taxonomy of those aspects of an extra-linguistic context that are known to influence language use. For example, these situational parameters describe distinctions between written and spoken usage, the relationship between addresser and addressee, and the purpose of the text. We begin by briefly defining some key terms that will be used throughout this paper. First, context of production and communicative situation refer to the non-linguistic attributes of the environment in which a corpus was created.

The world is facing a maternal health crisis. According to the World Health Organization, approximately 810 women die each day due to preventable causes related to pregnancy and childbirth. Two-thirds of these deaths occur in sub-Saharan Africa. In Rwanda, one of the leading causes of maternal mortality is infected Cesarean section wounds. An interdisciplinary team of doctors and researchers from MIT, Harvard University, and Partners in Health (PIH) in Rwanda have proposed a solution to address this problem.

Traditionally, an agent maximizes a reward defined according to the task to perform: it may be a score when the agent learns to solve a game or a distance function when the agent learns to reach a goal. The reward is then considered as extrinsic (or as a feedback) because the reward function is provided expertly and specifically for the task. With an extrinsic reward, many spectacular results have been obtained on Atari game [Bellemare et al. 2015] with the Deep Q-network (DQN) [Mnih et al. 2015] through the integration of deep learning to RL, leading to deep reinforcement learning (DRL). However, despite the recent improvements of DRL approaches, they turn out to be most of the time unsuccessful when the rewards are scattered in the environment, as the agent is then unable to learn the desired behavior for the targeted task [Francois-Lavet et al. 2018]. Moreover, the behaviors learned by the agent are hardly reusable, both within the same task and across many different tasks [Francois-Lavet et al. 2018]. It is difficult for an agent to generalize the learnt skills to make high-level decisions in the environment. For example, such skill could be go to the door using primitive actions consisting in moving in the four cardinal directions; or even to move forward controlling different joints of a humanoid robot like in the robotic simulator MuJoCo [Todorov et al. 2012]. On another side, unlike RL, developmental learning [Cangelosi and Schlesinger 2018; Oudeyer and Smith 2016; Piaget and Cook 1952] is based on the trend that babies, or more broadly organisms, acquire new skill while spontaneously exploring their environment [Barto 2013; Gopnik et al. 1999].

Answering complex queries over knowledge graphs (KG) is an important yet challenging task because of the KG incompleteness issue and cascading errors during reasoning. Recent query embedding (QE) approaches to embed the entities and relations in a KG and the first-order logic (FOL) queries into a low dimensional space, answering queries by dense similarity search. However, previous works mainly concentrate on the target answers, ignoring intermediate entities' usefulness, which is essential for relieving the cascading error problem in logical query answering. In addition, these methods are usually designed with their own geometric or distributional embeddings to handle logical operators like union, intersection, and negation, with the sacrifice of the accuracy of the basic operator - projection, and they could not absorb other embedding methods to their models. In this work, we propose a Neural and Symbolic Entangled framework (ENeSy) for complex query answering, which enables the neural and symbolic reasoning to enhance each other to alleviate the cascading error and KG incompleteness. The projection operator in ENeSy could be any embedding method with the capability of link prediction, and the other FOL operators are handled without parameters. With both neural and symbolic reasoning results contained, ENeSy answers queries in ensembles. ENeSy achieves the SOTA performance on several benchmarks, especially in the setting of the training model only with the link prediction task.

Large quantifies of online user activity data, such as weekly web search volumes, which co-evolve with the mutual influence of several queries and locations, serve as an important social sensor. It is an important task to accurately forecast the future activity by discovering latent interactions from such data, i.e., the ecosystems between each query and the flow of influences between each area. However, this is a difficult problem in terms of data quantity and complex patterns covering the dynamics. To tackle the problem, we propose FluxCube, which is an effective mining method that forecasts large collections of co-evolving online user activity and provides good interpretability. Our model is the expansion of a combination of two mathematical models: a reaction-diffusion system provides a framework for modeling the flow of influences between local area groups and an ecological system models the latent interactions between each query. Also, by leveraging the concept of physics-informed neural networks, FluxCube achieves high interpretability obtained from the parameters and high forecasting performance, together. Extensive experiments on real datasets showed that FluxCube outperforms comparable models in terms of the forecasting accuracy, and each component in FluxCube contributes to the enhanced performance. We then show some case studies that FluxCube can extract useful latent interactions between queries and area groups.

Quotation extraction aims to extract quotations from written text. There are three components in a quotation: source refers to the holder of the quotation, cue is the trigger word(s), and content is the main body. Existing solutions for quotation extraction mainly utilize rule-based approaches and sequence labeling models. While rule-based approaches often lead to low recalls, sequence labeling models cannot well handle quotations with complicated structures. In this paper, we propose the Context and Former-Label Enhanced Net (CofeNet) for quotation extraction. CofeNet is able to extract complicated quotations with components of variable lengths and complicated structures. On two public datasets (i.e., PolNeAR and Riqua) and one proprietary dataset (i.e., PoliticsZH), we show that our CofeNet achieves state-of-the-art performance on complicated quotation extraction.