This paper introduces a novel approach to vector graphics drawing by formalizing it within Category Theory (CT). Using the categorical programming paradigm, CT abstractions are interpreted as design patterns within Functional Programming (FP), bridging theoretical formalism and practical implementation. Building on Yorgey’s work, we represent diagrams as lists of graphical primitives, serving as a normal form for vector graphics. Diagrams are manipulated via graphical transformations (e.g., translation, rotation, scaling, color change) and list concatenation, which are modeled as an endofunctor F. This enables the construction of a free monad 𝕋≅ as a way to accumulate operations into expression trees lazily. We formalize graphical marks as abstractions over diagrams via slice categories, and vector graphics as values of the free monad 𝕋 over marks. Lastly, we introduce graphic expressions as a way to leverage F-algebras and F-coalgebras to construct complex graphics systematically. A prototype implementation in Lean demonstrates the applicability of our theory in a programming language based on dependent types, illustrating how categorical programming bridges the gap between mathematical abstraction and practical programming. An implementation in Julia is presented elsewhere.

This short paper describes the first steps in a project to construct a knowledge graph for Brazilian history based on the Brazilian Dictionary of Historical Biographies (DHBB) and Wikipedia/Wikidata. We contend that large repositories of Brazilian-named entities (people, places, organizations, and political events and movements) would be beneficial for extracting information from Portuguese texts. We show that many of the terms/entities described in the DHBB do not have corresponding concepts (or Q items) in Wikidata, the largest structured database of entities associated with Wikipedia. We describe previous work on extracting information from the DHBB and outline the steps to construct a Wikidata-based historical knowledge graph.

Semantic role labeling (SRL) identifies the predicate-argument structure in a sentence. This task is usually accomplished in four steps: predicate identification, predicate sense disambiguation, argument identification, and argument classification. Errors introduced at one step propagate to later steps. Unfortunately, the existing SRL evaluation scripts do not consider the full effect of this error propagation aspect. They either evaluate arguments independent of predicate sense (CoNLL09) or do not evaluate predicate sense at all (CoNLL05), yielding an inaccurate SRL model performance on the argument classification task. In this paper, we address key practical issues with existing evaluation scripts and propose a more strict SRL evaluation metric PriMeSRL. We observe that by employing PriMeSRL, the quality evaluation of all SoTA SRL models drops significantly, and their relative rankings also change. We also show that PriMeSRLsuccessfully penalizes actual failures in SoTA SRL models.

In 2008, the Princeton team released the last version of the “Princeton Annotated Gloss Corpus”. In this corpus. The word forms from the definitions and examples (glosses) of Princeton WordNet are manually linked to the context-appropriate sense in WordNet. However, the annotation was not complete, and the dataset was never officially released as part of WordNet 3.0, remaining as one of the standoff files available for download. Eleven years later, in 2019, one of the authors of this paper restarted the project aiming to complete the sense annotation of the approximately 200 thousand word forms not yet annotated. Here, we provide additional motivations to complete this dataset and report the progress in the work and evaluations. Intending to provide an extra level of consistency in the sense annotation and a deep semantic representation of the definitions and examples promoting WordNet from a lexical resource to a lightweight ontology, we now employ the English Resource Grammar (ERG), a broad-coverage HPSG grammar of English to parse the sentences and project the sense annotations from the surface words to the ERG predicates. We also report some initial steps on upgrading the corpus to WordNet 3.1 to facilitate mapping the data to other lexical resources.

In this project note we describe our work to make better documentation for the Open Multilingual Wordnet (OMW), a platform integrating many open wordnets. This includes the documentation of the OMW website itself as well as of semantic relations used by the component wordnets. Some of this documentation work was done with the support of the Google Season of Docs. The OMW project page, which links both to the actual OMW server and the documentation has been moved to a new location: https://omwn.org.

ULKB Logic is an open-source framework written in Python for reasoning over knowledge graphs. It provides an interactive theorem prover-like environment equipped with a higher-order language similar to the one used by HOL Light. The main goal of ULKB Logic is to ease the construction of applications that combine state-of-the-art computational logic tools with the knowledge available in knowledge graphs, such as Wikidata. To this end, the framework provides APIs for fetching statements from SPARQL endpoints and operating over the constructed theories using automated theorem provers and SMT solvers (such as the E prover and Z3). In this paper, we describe the design and implementation of ULKB Logic, its interfaces for querying knowledge graphs and calling external provers, and plans for further development.

We propose a methodology to populate the verb type hierarchy of a deep computational grammar in the HPSG formalism using syntactic and morphological information from Universal Dependencies (UD) treebanks. It is exemplified by means of the UD Bosque corpus and PorGram, a computational grammar for Portuguese constructed in the LinGO Grammar Matrix framework, but it can be applied to analogous grammars of other languages using other UD treebanks. The main component of the methodology is a Python module that extracts from the annotated sentences the core arguments and other features that are relevant to determine verb valence. This module enables the creation of a Python dictionary that maps valence frames to verb objects. This dictionary facilitates not only determining which frames occur with which verbs, but also detecting annotation errors. The potential of the module for rapid expansion of the lexical coverage of PorGram and corpus annotation error detection is illustrated with concrete examples.

Semantic role labeling (SRL) represents the meaning of a sentence in the form of predicate-argument structures. Such shallow semantic analysis is helpful in a wide range of downstream NLP tasks and real-world applications. As treebanks enabled the development of powerful syntactic parsers, high-quality training data in the form of propbanks is crucial to build models for accurate predicate-argument analysis. Unfortunately, most languages simply do not have corresponding propbanks due to the high cost required to construct such resources. To overcome such challenges, we released Universal Proposition Bank 1.0 (UP1.0) in 2017, with high-quality propbank data generated via a two-stage method exploiting monolingual SRL and multilingual parallel data. In this paper, we introduce Universal Proposition Bank 2.0 (UP2.0), with significant enhancements over UP1.0, including: (1) propbanks with higher quality by using a state-of-the-art monolingual SRL and improved auto-generation of annotations; (2) expanded language coverage (from 7 to 23 languages); (3) span annotation for the decoupling of syntactic analysis; and (4) gold data for a subset of the languages. We also share our experimental results that confirm the significant quality improvements of the generated propbanks. In addition, we present a comprehensive experimental evaluation on how different implementation choices impact the quality of the resulting data. We release these resources to the research community and hope to encourage more research on cross-lingual SRL.

A HPSG é uma teoria gramatical lexicalista que propõe a formalização em paralelo das estruturas morfossintáticas e semânticas. Este trabalho descreve a implementação computacional de valências verbais numa nova gramática do português nesse formalismo. Essa gramática é relevante não somente para aplicações de compreensão textual, mas representa também uma contribuição à documentação formal das estruturas da língua, destacando-se pelo tratamento das construções de controle e alçamento. A gramática tem sido implementada incrementalmente por meio da sua aplicação a conjuntos de teste cada vez mais abrangentes. Com 278 entradas e um total de 215 lemas, o léxico verbal ainda é pequeno. No entanto, a hierarquia de tipos proposta modela as propriedades de 118 classes valenciais, das quais 57 são tipos que codificam classes de verbos. A gramática analisa 94% de um total de 581 sentenças gramaticais, apresentando, ao mesmo tempo, baixa hipergeração em um conjunto de 167 exemplos agramaticais.

This paper investigates updates of Universal Dependencies (UD) treebanks in 23 languages and their impact on a downstream application. Numerous people are involved in updating UD’s annotation guidelines and treebanks in various languages. However, it is not easy to verify whether the updated resources maintain universality with other language resources. Thus, validity and consistency of multilingual corpora should be tested through application tasks involving syntactic structures with PoS tags, dependency labels, and universal features. We apply the syntactic parsers trained on UD treebanks from multiple versions (2.0 to 2.7) to a clause-level sentiment extractor. We then analyze the relationships between attachment scores of dependency parsers and performance in application tasks. For future UD developments, we show examples of outputs that differ depending on version.

Spatial reasoning is an important part of common-sense reasoning but has most often been examined in isolation from other areas of common sense knowledge representation. Spatial reasoning also has been a prominent area of research in linguistic semantics. In this work, we aim to create a corpus of common sense reasoning problems formalized in the context of a large and comprehensive theory of world knowledge, in the hopes that it will be more generally reusable on a wide variety of practical reasoning problems. An additional goal is to demonstrate the computational sufficiency of this work by employing automated reasoning to solve the problems. Finally, we also state the problems first in natural language, to provide a set of test cases for theories of linguistic semantics and computational linguistics, with a computational representation that can help to validate whether any method or process of interpreting language into a computational logic representation is correct or sufficient. This is an ambitious set of goals, and it will not be possible to provide solutions for all the problems that could be formulated in natural language. In particular, we do not attempt to create an implementation, at least at this time, for a system that can translate all of the stated problems from language to logic. Many of the problems also appear to require a logic and reasoning system beyond first order logic. It is important however, to have challenges that are unsolved in order to motivate research. We utilize the Suggested Upper Merged Ontology (SUMO)[5, 7], a comprehensive ontology of around 20,000 concepts and 80,000 hand-authored logical statements in a higher-order logic, that has an associated integrated development environment[9] integrated with leading theorem provers such as Eprover [10] Vampire [4] and LEO-II [1], and manually-created links[6] to the WordNet lexico-semantic database[3]. We described [9] elsewhere how to translate SUMO to the strictly first order language of TPTP [12], as well as TF0 [8] and THF[2].

Ampla cobertura lexical constitui um dos pré-requisitos para a robustez de uma gramática computacional. Propomos uma metodologia para povoar a tabela de formas verbais irregulares da PorGram (gramática do português no formalismo HPSG) utilizando os dados do dicionário MorphoBr. Nós implementamos um algoritmo que classifica as formas verbais do MorphoBr em regulares e irregulares, aplicando as regras flexionais da PorGram. Avaliamos o algoritmo com numa amostra de 38 verbos, regulares e irregulares, obtendo os resultados esperados. Uma contribuição adicional foi o melhoramento do MorphoBr, com a eliminação de mais de 270.000 entradas erradas e a inclusão de quase 13.000 entradas faltantes.

The Global Wordnet Formats have been introduced to enable wordnets to have a common representation that can be integrated through the Global WordNet Grid. As a result of their adoption, a number of shortcomings of the format were identified, and in this paper we describe the extensions to the formats that address these issues. These include: ordering of senses, dependencies between wordnets, pronunciation, syntactic modelling, relations, sense keys, metadata and RDF support. Furthermore, we provide some perspectives on how these changes help in the integration of wordnets.

Recursos linguísticos, como corpora, são fundamentais para o desenvolvimento de ferramentas para processamento de textos. No processamento de textos em português, um recurso atualmente considerado fundamental é o corpus UD Bosque, parte da coleção de corpora no projeto 'Universal Dependencies' (UD). A despeito do corpus UD Bosque ter sido convertido para as anotações de UD de um corpus originalmente revisado, ainda são vários os problemas de consistência das anotações encontrados na atual versão do corpus. Neste trabalho, apresentamos a metodologia usada para corrigir os problemas de anotações morfológicas nos corpus UD Bosque, em particular, identificamos erros nas anotações morfológicas de determinantes e adjetivos que deveriam concordar com os substantivos que modificam. Discutimos como os erros foram identificados, as exceções ou casos não triviais, as correções realizadas e o impacto das mudanças no corpus no treinamento de analisadores sintáticos estatísticos.

Ontologies are computational artifacts that model consensual aspects of reality. In distributed contexts, applications often need to utilize information from several distinct ontologies. In order to integrate multiple ontologies, entities modeled in each ontology must be matched through an ontology alignment. However, imperfect alignments may introduce inconsistencies. One kind of inconsistency, which is often introduced, is the violation of the conservativity principle, that states that the alignment should not introduce new subsumption relations between entities from the same source ontology. We propose a two-step quadratic-time algorithm for automatically correcting such violations, and evaluate it against datasets from the Ontology Alignment Evaluation Initiative 2019, comparing the results to a state-of-the-art approach. The proposed algorithm was significantly faster and less aggressive; that is, it performed fewer modifications over the original alignment when compared to the state-of-the-art algorithm.

The maintenance of wordnets and lexical knwoledge bases typically relies on time-consuming manual effort. In order to minimise this issue, we propose the exploitation of models of distributional semantics, namely word embeddings learned from corpora, in the automatic identification of relation instances missing in a wordnet. Analogy-solving methods are first used for learning a set of relations from analogy tests focused on each relation. Despite their low accuracy, we noted that a portion of the top-given answers are good suggestions of relation instances that could be included in the wordnet. This procedure is applied to the enrichment of OpenWordNet-PT, a public Portuguese wordnet. Relations are learned from data acquired from this resource, and illustrative examples are provided. Results are promising for accelerating the identification of missing relation instances, as we estimate that about 17% of the potential suggestions are good, a proportion that almost doubles if some are automatically invalidated.

This multilingual resource contains corpora for 14 languages, gathered at the occasion of the 1.2 edition of the PARSEME Shared Task on semi-supervised Identification of Verbal MWEs (2020). These corpora were meant to serve as additional "raw" corpora, to help discovering unseen verbal MWEs. The corpora are provided in CONLL-U (https://universaldependencies.org/format.html) format. They contain morphosyntactic annotations (parts of speech, lemmas, morphological features, and syntactic dependencies). Depending on the language, the information comes from treebanks (mostly Universal Dependencies v2.x) or from automatic parsers trained on UD v2.x treebanks (e.g., UDPipe). VMWEs include idioms (let the cat out of the bag), light-verb constructions (make a decision), verb-particle constructions (give up), inherently reflexive verbs (help oneself), and multi-verb constructions (make do). For the 1.2 shared task edition, the data covers 14 languages, for which VMWEs were annotated according to the universal guidelines. The corpora are provided in the cupt format, inspired by the CONLL-U format. Morphological and syntactic information ­­­­– not necessarily using UD tagsets – including parts of speech, lemmas, morphological features and/or syntactic dependencies are also provided. Depending on the language, the information comes from treebanks (e.g., Universal Dependencies) or from automatic parsers trained on treebanks (e.g., UDPipe). This item contains training, development and test data, as well as the evaluation tools used in the PARSEME Shared Task 1.2 (2020). The annotation guidelines are available online: http://parsemefr.lif.univ-mrs.fr/parseme-st-guidelines/1.2

We present our ongoing efforts towards the creation of a new Portuguese corpus based on the “Dicionário Histórico-Bibliográfico Brasileiro”. The aim to add as many linguistic annotations as possible using widely accepted annotation schemas and distributing all data in standard formats. This first exploratory work revisits what is already done and tests different tools to detect errors and look for the best methods to tackle the problem. Data is available at https://github.com/cpdoc/dhbb-nlp , and it will be continuously improved.