Search

In this contribution, we describe the system presented by the PolyU CBS-Comp Team at the Task 1 of SemEval 2021, where the goal was the estimation of the complexity of words in a given sentence context. Our top system, based on a combination of lexical, syntactic, word embeddings and Transformers-derived features and on a Gradient Boosting Regressor, achieves a top correlation score of 0.754 on the subtask 1 for single words and 0.659 on the subtask 2 for multiword expressions.

This paper reports a linguistically-enriched method of detecting token-level metaphors for the second shared task on Metaphor Detection. We participate in all four phases of competition with both datasets, i.e. Verbs and AllPOS on the VUA and the TOFEL datasets. We use the modality exclusivity and embodiment norms for constructing a conceptual representation of the nodes and the context. Our system obtains an F-score of 0.652 for the VUA Verbs track, which is 5% higher than the strong baselines. The experimental results across models and datasets indicate the salient contribution of using modality exclusivity and modality shift information for predicting metaphoricity.

In this paper, we present the findings of the Third VarDial Evaluation Campaign organized as part of the sixth edition of the workshop on Natural Language Processing (NLP) for Similar Languages, Varieties and Dialects (VarDial), co-located with NAACL 2019. This year, the campaign included five shared tasks, including one task re-run – German Dialect Identification (GDI) – and four new tasks – Cross-lingual Morphological Analysis (CMA), Discriminating between Mainland and Taiwan variation of Mandarin Chinese (DMT), Moldavian vs. Romanian Cross-dialect Topic identification (MRC), and Cuneiform Language Identification (CLI). A total of 22 teams submitted runs across the five shared tasks. After the end of the competition, we received 14 system description papers, which are published in the VarDial workshop proceedings and referred to in this report.

Attention models are proposed in sentiment analysis because some words are more important than others. However,most existing methods either use local context based text information or user preference information. In this work, we propose a novel attention model trained by cognition grounded eye-tracking data. A reading prediction model is first built using eye-tracking data as dependent data and other features in the context as independent data. The predicted reading time is then used to build a cognition based attention (CBA) layer for neural sentiment analysis. As a comprehensive model, We can capture attentions of words in sentences as well as sentences in documents. Different attention mechanisms can also be incorporated to capture other aspects of attentions. Evaluations show the CBA based method outperforms the state-of-the-art local context based attention methods significantly. This brings insight to how cognition grounded data can be brought into NLP tasks.

Metaphor detection has been both challenging and rewarding in natural language processing applications. This study offers a new approach based on eventive information in detecting metaphors by leveraging the Chinese writing system, which is a culturally bound ontological system organized according to the basic concepts represented by radicals. As such, the information represented is available in all Chinese text without pre-processing. Since metaphor detection is another culturally based conceptual representation, we hypothesize that sub-textual information can facilitate the identification and classification of the types of metaphoric events denoted in Chinese text. We propose a set of syntactic conditions crucial to event structures to improve the model based on the classification of radical groups. With the proposed syntactic conditions, the model achieves a performance of 0.8859 in terms of F-scores, making 1.7% of improvement than the same classifier with only Bag-of-word features. Results show that eventive information can improve the effectiveness of metaphor detection. Event information is rooted in every language, and thus this approach has a high potential to be applied to metaphor detection in other languages.

Several studies on sentence processing suggest that the mental lexicon keeps track of the mutual expectations between words. Current DSMs, however, represent context words as separate features, thereby loosing important information for word expectations, such as word interrelations. In this paper, we present a DSM that addresses this issue by defining verb contexts as joint syntactic dependencies. We test our representation in a verb similarity task on two datasets, showing that joint contexts achieve performances comparable to single dependencies or even better. Moreover, they are able to overcome the data sparsity problem of joint feature spaces, in spite of the limited size of our training corpus., 5 pages

The manual Chinese word segmentation dataset WordSegCHC 1.0 which was built by eight crowdsourcing tasks conducted on the Crowdflower platform contains the manual word segmentation data of 152 Chinese sentences whose length ranges from 20 to 46 characters without punctuations. All the sentences received 200 segmentation responses in their corresponding crowdsourcing tasks and the numbers of valid response of them range from 123 to 143 (each sentence was segmented by more than 120 subjects). We also proposed an evaluation method called manual segmentation error rate (MSER) to evaluate the dataset; the MSER of the dataset is proved to be very low which indicates reliable data quality. In this work, we applied the crowdsourcing method to Chinese word segmentation task and the results confirmed again that the crowdsourcing method is a promising tool for linguistic data collection; the framework of crowdsourcing linguistic data collection used in this work can be reused in similar tasks; the resultant dataset filled a gap in Chinese language resources to the best of our knowledge, and it has potential applications in the research of word intuition of Chinese speakers and Chinese language processing.

In this paper, we describe the system we built for Task 11 of SemEval2015, which aims at identifying the sentiment intensity of figurative language in tweets. We use various features, including those specially concerned with the identification of irony and sarcasm. The features are evaluated through a decision tree regression model and a support vector regression model. The experiment result of the fivecross validation on the training data shows that the tree regression model outperforms the support vector regression model. The former is therefore used for the final evaluation of the task. The results show that our model performs especially well in predicting the sentiment intensity of tweets involving irony and sarcasm.

In this paper, we introduce EVALution 1.0, a dataset designed for the training and the evaluation of Distributional Semantic Models (DSMs). This version consists of almost 7.5K tuples, instantiating several semantic relations between word pairs (including hypernymy, synonymy, antonymy, meronymy). The dataset is enriched with a large amount of additional information (i.e. relation domain, word frequency, word POS, word semantic field, etc.) that can be used for either filtering the pairs or performing an in-depth analysis of the results. The tuples were extracted from a combination of ConceptNet 5.0 and WordNet 4.0, and subsequently filtered through automatic methods and crowdsourcing in order to ensure their quality. The dataset is freely downloadable1. An extension in RDF format, including also scripts for data processing, is under development.

The synergy between language sciences and language technology has been an elusive one for the computational linguistics community, especially when dealing with a language other than English. The reasons are two-fold: the lack of an accessible comprehensive and robust account of a specific language so as to allow strategic linking between a processing task to linguistic devices, and the lack of successful computational studies taking advantage of such links. With a fast growing number of available online resources, as well as a rapidly increasing number of members of the CL community who are interested in and/or working on Chinese language processing, the time is ripe to take a serious look at how knowledge of Chinese can help Chinese language processing.

Lexical Markup Framework (LMF, ISO-24613) is the ISO standard which provides a common standardized framework for the construction of natural language processing lexicons. LMF facilitates data exchange among computational linguistic resources, and also promises a convenient uniformity for future application. This study describes the design and implementation of the WordNet-LMF used to represent lexical semantics in Chinese WordNet. The compiled CWN-LMF will be released to the community for linguistic researches.

In text categorization, feature selection (FS) is a strategy that aims at making text classifiers more efficient and accurate. However, when dealing with a new task, it is still difficult to quickly select a suitable one from various FS methods provided by many previous studies. In this paper, we propose a theoretic framework of FS methods based on two basic measurements: frequency measurement and ratio measurement. Then six popular FS methods are in detail discussed under this framework. Moreover, with the guidance of our theoretical analysis, we propose a novel method called weighed frequency and odds (WFO) that combines the two measurements with trained weights. The experimental results on data sets from both topic-based and sentiment classification tasks show that this new method is robust across different tasks and numbers of selected features.

This paper reports prototype multilingual query expansion system relying on LMF compliant lexical resources. The system is one of the deliverables of a three-year project aiming at establishing an international standard for language resources which is applicable to Asian languages. Our important contributions to ISO 24613, standard Lexical Markup Framework (LMF) include its robustness to deal with Asian languages, and its applicability to cross-lingual query tasks, as illustrated by the prototype introduced in this paper.

In this paper we propose a model for conceptual access to multilingual lexicon based on shared orthography. Our proposal relies crucially on two facts: That both Chinese and Japanese conventionally use Chinese orthography in their respective writing systems, and that the Chinese orthography is anchored on a system of radical parts which encodes basic concepts. Each orthographic unit, called hanzi and kanji respectively, contains a radical which indicates the broad semantic class of the meaning of that unit. Our study utilizes the homomorphism between the Chinese hanzi and Japanese kanji systems to identify bilingual word correspondences. We use bilingual dictionaries, including WordNet, to verify semantic relation between the crosslingual pairs. These bilingual pairs are then mapped to an ontology constructed based on relations to the relation between the meaning of each character and the basic concept of their radical parts. The conceptual structure of the radical ontology is proposed as a model for simultaneous conceptual access to both languages. A study based on words containing characters composed of the "(mouth)" radical is given to illustrate the proposal and the actual model. The fact that this model works for two typologically very different languages and that the model contains generative lexicon like coersive links suggests that this model has the conceptual robustness to be applied to other languages.

Identification of transliterated names is a particularly difficult task of Named Entity Recognition (NER), especially in the Chinese context. Of all possible variations of transliterated named entities, the difference between PRC and Taiwan is the most prevalent and most challenging. In this paper, we introduce a novel approach to the automatic extraction of diverging transliterations of foreign named entities by bootstrapping co-occurrence statistics from tagged and segmented Chinese corpus. Preliminary experiment yields promising results and shows its potential in NLP applications.

This paper addresses two remaining challenges in Chinese word segmentation. The challenge in HLT is to find a robust segmentation method that requires no prior lexical knowledge and no extensive training to adapt to new types of data. The challenge in modelling human cognition and acquisition it to segment words efficiently without using knowledge of wordhood. We propose a radical method of word segmentation to meet both challenges. The most critical concept that we introduce is that Chinese word segmentation is the classification of a string of character-boundaries (CB's) into either word-boundaries (WB's) and non-word-boundaries. In Chinese, CB's are delimited and distributed in between two characters. Hence we can use the distributional properties of CB among the background character strings to predict which CB's are WB's.

In this paper we present an application fostering the integration and interoperability of computational lexicons, focusing on the particular case of mutual linking and cross-lingual enrichment of two wordnets, the ItalWordNet and Sinica BOW lexicons. This is intended as a case-study investigating the needs and requirements of semi-automatic integration and interoperability of lexical resources.

This paper describes an on-going project concerning with an ontological lexical resource based on the abundant conceptual information grounded on Chinese characters. The ultimate goal of this project is set to construct a cognitively sound and computationally effective character-grounded machine-understandable resource. Philosophically, Chinese ideogram has its ontological status, but its applicability to the NLP task has not been expressed explicitly in terms of language resource. We thus propose the first attempt to locate Chinese characters within the context of ontology. Having the primary success in applying it to some NLP tasks, we believe that the construction of this knowledge resource will shed new light on theoretical setting as well as the construction of Chinese lexical semantic resources.

The goal of this paper is to integrate the Conceptual Mapping Model with an ontology-based knowledge representation (i.e. Suggested Upper Merged Ontology (SUMO)) in order to demonstrate that conceptual metaphor analysis can be restricted and eventually, automated. In particular, we will propose a corpora-based operational definition for Mapping Principles, which are explanations of why a conventional conceptual metaphor has a particular source-target domain pairing. This paper will examine 2000 random examples of 'economy' (jingji) in Mandarin Chinese and postulate Mapping Principles based frequency and delimited with SUMO.

OLACMS (stands for Open Language Archives Community Metadata Set) is a standard for describe language resources. This paper provides suggestion to OLACMS 0.4 version by comparing it with other standards and applying it to Chinese and Formosan languages.

We present in this paper a series of induced methods to assign domain tags to WordNet entries. Our prime objective is to enrich the contextual information in WordNet specific to each synset entry. By using the available lexical sources such as Far East Dictionary and the contextual information in WordNet itself, we can find a foundation upon which we can base our categorization. Next we further examine the similarity between common lexical taxonomy and the semantic hierarchy of WordNet. Based on this observation and the knowledge of other semantic relations we enlarge the coverage of our findings in a systematic way. Evaluation of the results shows that we achieved reasonable and satisfactory accuracy. We propose this as the first step of wordnet expansion into a bona fide semantic network linked to real-world knowledge.

Establishing correspondences between wordnets of different languages is essential to both multilingual knowledge processing and for bootstrapping wordnets of low-density languages. We claim that such correspondences must be based on lexical semantic relations, rather than top ontology or word translations. In particular, we define a translation equivalence relation as a bilingual lexical semantic relation. Such relations can then be part of a logical entailment predicting whether source language semantic relations will hold in a target language or not. Our claim is tested with a study of 210 Chinese lexical lemmas and their possible semantic relations links bootstrapped from the Princeton WordNet. The results show that lexical semantic relation translations are indeed highly precise when they are logically inferable.

Assignment of grammatical categories is the fundamental step in natural language processing. And ambiguity resolution is one of the most challenging NLP tasks that is currently still beyond the power of machines. When two questions are combined together, the problem of resolution of categorical ambiguity is what a computational linguistic system can do reasonably good, but yet still unable to mimic the excellence of human beings. This task is even more challenging in Chinese language processing because of the poverty of morphological information to mark categories and the lack of convention to mark word boundaries. In this paper, we try to investigate the nature of categorical ambiguity in Chinese based on Sinica Corpus. The study differs crucially from previous studies in that it directly measure information content as the degree of ambiguity. This method not only offers an alternative interpretation of ambiguity, it also allows a different measure of success of categorical disambiguation. Instead of precision or recall, we can also measure by how much the information load has been reduced. This approach also allows us to identify which are the most ambiguous words in terms of information content. The somewhat surprising result actually reinforces the Saussurian view that underlying the systemic linguistic structure, assignment of linguistic content for each linguistic symbol is arbitrary.

This paper describes the design criteria and annotation guidelines of Sinica Treebank. The three design criteria are: Maximal Resource Sharing, Minimal Structural Complexity, and Optimal Semantic Information. One of the important design decisions following these criteria is the encoding of thematic role information. An on-line interface facilitating empirical studies of Chinese phrase structure is also described.

This paper proposes a segmentation standard for Chinese natural language processing. The standard is proposed to achieve linguistic felicity, computational feasibility, and data uniformity. Linguistic felicity is maintained by defining a segmentation unit to be equivalent to the theoretical definition of word, and by providing a set of segmentation principles that are equivalent to a functional definition of a word. Computational feasibility is ensured by the fact that the above functional definitions are procedural in nature and can be converted to segmentation algorithms, as well as by the implementable heuristic guidelines which deal with specific linguistic categories. Data uniformity is achieved by stratification of the standard itself and by defining a standard lexicon as part of the segmentation standard.

This is a project note on the first stage of the construction of a comprehensive corpus of both Modern and Classical Chinese. The corpus is built with the dual aim of serving as the central database for Chinese language processing and for supporting in-depth linguistic research in Mandarin Chinese.

In this paper we propose a framework of Information-based Case Grammar (ICG). This grammatical formalism entails that the lexical entry for each word contain both semantic and syntactic feature structures. In the feature structure of a phrasal head, we encode syntactic and semantic constraints on grammatical phrasal patterns in terms of thematic structures, and encode the precedence relations in terms of adjunct structures. Such feature structures denote partial information which defines the set of legal phrases. They also provide sufficient information to identify thematic roles. With this formalism, parsing and thematic analysis can be achieved simultaneously. Due to the simplicity and flexibility of Information-based Case Grammar, context dependent and discontinuous relations such as agreements, coordinations, long-distance dependencies, and control and binding, can be easily expressed. ICG is a kind of unification-based formalism. Therefore it inherits the advantages of unification-based formalisms and more.