Four years on from two papers (Belz et al., 2020; Howcroft et al., 2020) that first called out the lack of standardisation and comparability in the quality criteria assessed in NLP system evaluations, researchers still use widely differing quality criteria names and definitions, meaning that it continues to be unclear when the same aspect of quality is being assessed in two evaluations. While normalised quality criteria were proposed at the time, the list was unwieldy and using it came with a steep learning curve. In this demo paper, our aim is to address these issues with an interactive taxonomy tool that enables quick perusal and selection of the quality criteria, and provides decision support and examples of use at each node.
Wikipedia is known to have systematic gaps in its coverage that correspond to under-resourced languages as well as underrepresented groups. This paper presents a new tool to support efforts to fill in these gaps by automatically generating draft articles and facilitating post-editing and uploading to Wikipedia. A rule-based generator and an input-constrained LLM are used to generate two alternative articles, enabling the often more fluent, but error-prone, LLM-generated article to be content-checked against the more reliable, but less fluent, rule-generated article.
Following numerous calls in the literature for improved practices and standardisation in human evaluation in Natural Language Processing over the past ten years, we held a tutorial on the topic at the 2024 INLG Conference. The tutorial addressed the structure, development, design, implementation, execution and analysis of human evaluations of NLP system quality. Hands-on practical sessions were run, designed to facilitate assimilation of the material presented. Slides, lecture recordings, code and data have been made available on GitHub (https://github.com/Human-Evaluation-Tutorial/INLG-2024-Tutorial). In this paper, we provide summaries of the content of the eight units of the tutorial, alongside its research context and aims.
We present an overview of the GEM 2024 shared task, which comprised of both data-to-text generation and summarization. New datasets were compiled specifically for the task to reduce data contamination in the large language models, which the participants were likely to use. The paper describes the tasks, the datasets, the participating systems, the evaluation methods, and some preliminary results. The full results will be presented at INLG ‘24.
Our submission to the GEM data-to-text shared task aims to assess the quality of texts produced by the combination of a rule-based system with a language model of reduced size, by first using a rule-based generator to convert input triples into semantically correct English text, and then a language model to paraphrase these texts to make them more fluent. The texts are translated to languages other than English with the NLLB machine translation system.
At the heart of the Pyramid evaluation method for text summarization lie human written summary content units (SCUs). These SCUs areconcise sentences that decompose a summary into small facts. Such SCUs can be used to judge the quality of a candidate summary, possibly partially automated via natural language inference (NLI) systems. Interestingly, with the aim to fully automate the Pyramid evaluation, Zhang and Bansal (2021) show that SCUs can be approximated by automatically generated semantic role triplets (STUs). However, several questions currently lack answers, in particular: i) Are there other ways of approximating SCUs that can offer advantages?ii) Under which conditions are SCUs (or their approximations) offering the most value? In this work, we examine two novel strategiesto approximate SCUs: generating SCU approximations from AMR meaning representations (SMUs) and from large language models (SGUs), respectively. We find that while STUs and SMUs are competitive, the best approximation quality is achieved by SGUs. We also show through a simple sentence-decomposition baseline (SSUs) that SCUs (and their approximations) offer the most value when rankingshort summaries, but may not help as much when ranking systems or longer summaries.
To prevent the costly and inefficient use of resources on low-quality annotations, we want a method for creating a pool of dependable annotators who can effectively complete difficult tasks, such as evaluating automatic summarization. Thus, we investigate the recruitment of high-quality Amazon Mechanical Turk workers via a two-step pipeline. We show that we can successfully filter out subpar workers before they carry out the evaluations and obtain high-agreement annotations with similar constraints on resources. Although our workers demonstrate a strong consensus among themselves and CloudResearch workers, their alignment with expert judgments on a subset of the data is not as expected and needs further training in correctness. This paper still serves as a best practice for the recruitment of qualified annotators in other challenging annotation tasks.
Human evaluation is widely regarded as the litmus test of quality in NLP. A basic requirementof all evaluations, but in particular where they are used for meta-evaluation, is that they should support the same conclusions if repeated. However, the reproducibility of human evaluations is virtually never queried, let alone formally tested, in NLP which means that their repeatability and the reproducibility of their results is currently an open question. This focused contribution reports our review of human evaluation experiments reported in NLP papers over the past five years which we assessed in terms oftheir ability to be rerun. Overall, we estimatethat just 5% of human evaluations are repeatable in the sense that (i) there are no prohibitivebarriers to repetition, and (ii) sufficient information about experimental design is publicly available for rerunning them. Our estimate goesup to about 20% when author help is sought. We complement this investigation with a survey of results concerning the reproducibilityof human evaluations where those are repeatable in the first place. Here we find worryinglylow degrees of reproducibility, both in terms ofsimilarity of scores and of findings supportedby them. We summarise what insights can begleaned so far regarding how to make humanevaluations in NLP more repeatable and morereproducible.
Data augmentation is an important method for evaluating the robustness of and enhancing the diversity of training data for natural language processing (NLP) models. In this paper, we present NL-Augmenter, a new participatory Python-based natural language (NL) augmentation framework which supports the creation of transformations (modifications to the data) and filters (data splits according to specific features). We describe the framework and an initial set of 117 transformations and 23 filters for a variety of NL tasks annotated with noisy descriptive tags. The transformations incorporate noise, intentional and accidental human mistakes, socio-linguistic variation, semantically-valid style, syntax changes, as well as artificial constructs that are unambiguous to humans. We demonstrate the efficacy of NL-Augmenter by using its transformations to analyze the robustness of popular language models. We find different models to be differently challenged on different tasks, with quasi-systematic score decreases. The infrastructure, datacards, and robustness evaluation results are publicly available on GitHub for the benefit of researchers working on paraphrase generation, robustness analysis, and low-resource NLP.
In this paper, we describe M-FleNS, a multilingual flexible plug-and-play architecture designed to accommodate neural and symbolic modules, and initially instantiated with rule-based modules. We focus on using M-FleNS for the specific purpose of building new resources for Irish, a language currently under-represented in the NLP landscape. We present the general M-FleNS framework and how we use it to build an Irish Natural Language Generation system for verbalising part of the DBpedia ontology and building a multilayered dataset with rich linguistic annotations. Via automatic and human assessments of the output texts we show that with very limited resources we are able to create a system that reaches high levels of fluency and semantic accuracy, while having very low energy and memory requirements.
Rule-based text generators lack the coverage and fluency of their neural counterparts, but have two big advantages over them: (i) they are entirely controllable and do not hallucinate; and (ii) they can fully explain how an output was generated from an input. In this paper we leverage these two advantages to create large and reliable synthetic datasets with multiple human-intelligible intermediate representations. We present the Modular Data-to-Text (Mod-D2T) Dataset which incorporates ten intermediate-level representations between input triple sets and output text; the mappings from one level to the next can broadly be interpreted as the traditional modular tasks of an NLG pipeline. We describe the Mod-D2T dataset, evaluate its quality via manual validation and discuss its applications and limitations. Data, code and documentation are available at https://github.com/mille-s/Mod-D2T.
We present work in progress that aims to address the coverage issue faced by rule-based text generators. We propose a pipeline for extracting abstract dependency template (predicate-argument structures) from Wikipedia text to be used as input for generating text from structured data with the FORGe system. The pipeline comprises three main components: (i) candidate sentence retrieval, (ii) clause extraction, ranking and selection, and (iii) conversion to predicate-argument form. We present an approach and preliminary evaluation for the ranking and selection module.
In this paper, we describe the submission of Dublin City University (DCU) and Trinity College Dublin (TCD) for the WebNLG 2023 shared task. We present a fully rule-based pipeline for generating Irish texts from DBpedia triple sets which comprises 4 components: triple lexicalisation, generation of noninflected Irish text, inflection generation, and post-processing.
This paper describes and tests a method for carrying out quantified reproducibility assessment (QRA) that is based on concepts and definitions from metrology. QRA produces a single score estimating the degree of reproducibility of a given system and evaluation measure, on the basis of the scores from, and differences between, different reproductions. We test QRA on 18 different system and evaluation measure combinations (involving diverse NLP tasks and types of evaluation), for each of which we have the original results and one to seven reproduction results. The proposed QRA method produces degree-of-reproducibility scores that are comparable across multiple reproductions not only of the same, but also of different, original studies. We find that the proposed method facilitates insights into causes of variation between reproductions, and as a result, allows conclusions to be drawn about what aspects of system and/or evaluation design need to be changed in order to improve reproducibility.
Evaluations in machine learning rarely use the latest metrics, datasets, or human evaluation in favor of remaining compatible with prior work. The compatibility, often facilitated through leaderboards, thus leads to outdated but standardized evaluation practices. We pose that the standardization is taking place in the wrong spot. Evaluation infrastructure should enable researchers to use the latest methods and what should be standardized instead is how to incorporate these new evaluation advances. We introduce GEMv2, the new version of the Generation, Evaluation, and Metrics Benchmark which uses a modular infrastructure for dataset, model, and metric developers to benefit from each other’s work. GEMv2 supports 40 documented datasets in 51 languages, ongoing online evaluation for all datasets, and our interactive tools make it easier to add new datasets to the living benchmark.
We introduce GEM, a living benchmark for natural language Generation (NLG), its Evaluation, and Metrics. Measuring progress in NLG relies on a constantly evolving ecosystem of automated metrics, datasets, and human evaluation standards. Due to this moving target, new models often still evaluate on divergent anglo-centric corpora with well-established, but flawed, metrics. This disconnect makes it challenging to identify the limitations of current models and opportunities for progress. Addressing this limitation, GEM provides an environment in which models can easily be applied to a wide set of tasks and in which evaluation strategies can be tested. Regular updates to the benchmark will help NLG research become more multilingual and evolve the challenge alongside models. This paper serves as the description of the data for the 2021 shared task at the associated GEM Workshop.
We present our Charles-UPF submission for the Shared Task on Evaluating Accuracy in Generated Texts at INLG 2021. Our system can detect the errors automatically using a combination of a rule-based natural language generation (NLG) system and pretrained language models (LMs). We first utilize a rule-based NLG system to generate sentences with facts that can be derived from the input. For each sentence we evaluate, we select a subset of facts which are relevant by measuring semantic similarity to the sentence in question. Finally, we finetune a pretrained language model on annotated data along with the relevant facts for fine-grained error detection. On the test set, we achieve 69% recall and 75% precision with a model trained on a mixture of human-annotated and synthetic data.
This paper reports results from a reproduction study in which we repeated the human evaluation of the PASS Dutch-language football report generation system (van der Lee et al., 2017). The work was carried out as part of the ReproGen Shared Task on Reproducibility of Human Evaluations in NLG, in Track A (Paper 1). We aimed to repeat the original study exactly, with the main difference that a different set of evaluators was used. We describe the study design, present the results from the original and the reproduction study, and then compare and analyse the differences between the two sets of results. For the two ‘headline’ results of average Fluency and Clarity, we find that in both studies, the system was rated more highly for Clarity than for Fluency, and Clarity had higher standard deviation. Clarity and Fluency ratings were higher, and their standard deviations lower, in the reproduction study than in the original study by substantial margins. Clarity had a higher degree of reproducibility than Fluency, as measured by the coefficient of variation. Data and code are publicly available.
In this paper, we present a pipeline system that generates architectural landmark descriptions using textual, visual and structured data. The pipeline comprises five main components:(i) a textual analysis component, which extracts information from Wikipedia pages; (ii)a visual analysis component, which extracts information from copyright-free images; (iii) a retrieval component, which gathers relevant (property, subject, object) triples from DBpedia; (iv) a fusion component, which stores the contents from the different modalities in a Knowledge Base (KB) and resolves the conflicts that stem from using different sources of information; (v) an NLG component, which verbalises the resulting contents of the KB. We show that thanks to the addition of other modalities, we can make the verbalisation of DBpedia triples more relevant and/or inspirational.
The RDF-to-text task has recently gained substantial attention due to the continuous growth of RDF knowledge graphs in number and size. Recent studies have focused on systematically comparing RDF-to-text approaches on benchmarking datasets such as WebNLG. Although some evaluation tools have already been proposed for text generation, none of the existing solutions abides by the Findability, Accessibility, Interoperability, and Reusability (FAIR) principles and involves RDF data for the knowledge extraction task. In this paper, we present BENG, a FAIR benchmarking platform for Natural Language Generation (NLG) and Knowledge Extraction systems with focus on RDF data. BENG builds upon the successful benchmarking platform GERBIL, is opensource and is publicly available along with the data it contains.
WebNLG+ offers two challenges: (i) mapping sets of RDF triples to English or Russian text (generation) and (ii) converting English or Russian text to sets of RDF triples (semantic parsing). Compared to the eponymous WebNLG challenge, WebNLG+ provides an extended dataset that enable the training, evaluation, and comparison of microplanners and semantic parsers. In this paper, we present the results of the generation and semantic parsing task for both English and Russian and provide a brief description of the participating systems.
Human assessment remains the most trusted form of evaluation in NLG, but highly diverse approaches and a proliferation of different quality criteria used by researchers make it difficult to compare results and draw conclusions across papers, with adverse implications for meta-evaluation and reproducibility. In this paper, we present (i) our dataset of 165 NLG papers with human evaluations, (ii) the annotation scheme we developed to label the papers for different aspects of evaluations, (iii) quantitative analyses of the annotations, and (iv) a set of recommendations for improving standards in evaluation reporting. We use the annotations as a basis for examining information included in evaluation reports, and levels of consistency in approaches, experimental design and terminology, focusing in particular on the 200+ different terms that have been used for evaluated aspects of quality. We conclude that due to a pervasive lack of clarity in reports and extreme diversity in approaches, human evaluation in NLG presents as extremely confused in 2020, and that the field is in urgent need of standard methods and terminology.
Current standards for designing and reporting human evaluations in NLP mean it is generally unclear which evaluations are comparable and can be expected to yield similar results when applied to the same system outputs. This has serious implications for reproducibility testing and meta-evaluation, in particular given that human evaluation is considered the gold standard against which the trustworthiness of automatic metrics is gauged. %and merging others, as well as deciding which evaluations should be able to reproduce each other’s results. Using examples from NLG, we propose a classification system for evaluations based on disentangling (i) what is being evaluated (which aspect of quality), and (ii) how it is evaluated in specific (a) evaluation modes and (b) experimental designs. We show that this approach provides a basis for determining comparability, hence for comparison of evaluations across papers, meta-evaluation experiments, reproducibility testing.
This paper presents results from the Third Shared Task on Multilingual Surface Realisation (SR’20) which was organised as part of the COLING’20 Workshop on Multilingual Surface Realisation. As in SR’18 and SR’19, the shared task comprised two tracks: (1) a Shallow Track where the inputs were full UD structures with word order information removed and tokens lemmatised; and (2) a Deep Track where additionally, functional words and morphological information were removed. Moreover, each track had two subtracks: (a) restricted-resource, where only the data provided or approved as part of a track could be used for training models, and (b) open-resource, where any data could be used. The Shallow Track was offered in 11 languages, whereas the Deep Track in 3 ones. Systems were evaluated using both automatic metrics and direct assessment by human evaluators in terms of Readability and Meaning Similarity to reference outputs. We present the evaluation results, along with descriptions of the SR’19 tracks, data and evaluation methods, as well as brief summaries of the participating systems. For full descriptions of the participating systems, please see the separate system reports elsewhere in this volume.
We report results from the SR’19 Shared Task, the second edition of a multilingual surface realisation task organised as part of the EMNLP’19 Workshop on Multilingual Surface Realisation. As in SR’18, the shared task comprised two tracks with different levels of complexity: (a) a shallow track where the inputs were full UD structures with word order information removed and tokens lemmatised; and (b) a deep track where additionally, functional words and morphological information were removed. The shallow track was offered in eleven, and the deep track in three languages. Systems were evaluated (a) automatically, using a range of intrinsic metrics, and (b) by human judges in terms of readability and meaning similarity. This report presents the evaluation results, along with descriptions of the SR’19 tracks, data and evaluation methods. For full descriptions of the participating systems, please see the separate system reports elsewhere in this volume.
This paper presents an exploratory study that aims to evaluate the usefulness of back-translation in Natural Language Generation (NLG) from semantic representations for non-English languages. Specifically, Abstract Meaning Representation and Brazilian Portuguese (BP) are chosen as semantic representation and language, respectively. Two methods (focused on Statistical and Neural Machine Translation) are evaluated on two datasets (one automatically generated and another one human-generated) to compare the performance in a real context. Also, several cuts according to quality measures are performed to evaluate the importance (or not) of the data quality in NLG. Results show that there are still many improvements to be made but this is a promising approach.
Statistical generators increasingly dominate the research in NLG. However, grammar-based generators that are grounded in a solid linguistic framework remain very competitive, especially for generation from deep knowledge structures. Furthermore, if built modularly, they can be ported to other genres and languages with a limited amount of work, without the need of the annotation of a considerable amount of training data. One of these generators is FORGe, which is based on the Meaning-Text Model. In the recent WebNLG challenge (the first comprehensive task addressing the mapping of RDF triples to text) FORGe ranked first with respect to the overall quality in human evaluation. We extend the coverage of FORGE’s open source grammatical and lexical resources for English, so as to further improve the English texts, and port them to Spanish, to achieve a comparable quality. This confirms that, as already observed in the case of SimpleNLG, a robust universal grammar-driven framework and a systematic organization of the linguistic resources can be an adequate choice for NLG applications.
We report results from the SR’18 Shared Task, a new multilingual surface realisation task organised as part of the ACL’18 Workshop on Multilingual Surface Realisation. As in its English-only predecessor task SR’11, the shared task comprised two tracks with different levels of complexity: (a) a shallow track where the inputs were full UD structures with word order information removed and tokens lemmatised; and (b) a deep track where additionally, functional words and morphological information were removed. The shallow track was offered in ten, and the deep track in three languages. Systems were evaluated (a) automatically, using a range of intrinsic metrics, and (b) by human judges in terms of readability and meaning similarity. This report presents the evaluation results, along with descriptions of the SR’18 tracks, data and evaluation methods. For full descriptions of the participating systems, please see the separate system reports elsewhere in this volume.
In this paper, we present the datasets used in the Shallow and Deep Tracks of the First Multilingual Surface Realisation Shared Task (SR’18). For the Shallow Track, data in ten languages has been released: Arabic, Czech, Dutch, English, Finnish, French, Italian, Portuguese, Russian and Spanish. For the Deep Track, data in three languages is made available: English, French and Spanish. We describe in detail how the datasets were derived from the Universal Dependencies V2.0, and report on an evaluation of the Deep Track input quality. In addition, we examine the motivation for, and likely usefulness of, deriving NLG inputs from annotations in resources originally developed for Natural Language Understanding (NLU), and assess whether the resulting inputs supply enough information of the right kind for the final stage in the NLG process.
An increasing amount of research tackles the challenge of text generation from abstract ontological or semantic structures, which are in their very nature potentially large connected graphs. These graphs must be “packaged” into sentence-wise subgraphs. We interpret the problem of sentence packaging as a community detection problem with post optimization. Experiments on the texts of the VerbNet/FrameNet structure annotated-Penn Treebank, which have been converted into graphs by a coreference merge using Stanford CoreNLP, show a high F1-score of 0.738.
We present the contribution of Universitat Pompeu Fabra’s NLP group to the SemEval Task 9.2 (AMR-to-English Generation). The proposed generation pipeline comprises: (i) a series of rule-based graph-transducers for the syntacticization of the input graphs and the resolution of morphological agreements, and (ii) an off-the-shelf statistical linearization component.
We propose a shared task on multilingual Surface Realization, i.e., on mapping unordered and uninflected universal dependency trees to correctly ordered and inflected sentences in a number of languages. A second deeper input will be available in which, in addition, functional words, fine-grained PoS and morphological information will be removed from the input trees. The first shared task on Surface Realization was carried out in 2011 with a similar setup, with a focus on English. We think that it is time for relaunching such a shared task effort in view of the arrival of Universal Dependencies annotated treebanks for a large number of languages on the one hand, and the increasing dominance of Deep Learning, which proved to be a game changer for NLP, on the other hand.
This demo paper presents the multilingual deep sentence generator developed by the TALN group at Universitat Pompeu Fabra, implemented as a series of rule-based graph-transducers for the syntacticization of the input graphs, the resolution of morphological agreements, and the linearization of the trees.
In this paper we describe the development of a text simplification system for Spanish. Text simplification is the adaptation of a text to the special needs of certain groups of readers, such as language learners, people with cognitive difficulties and elderly people, among others. There is a clear need for simplified texts, but manual production and adaptation of existing texts is labour intensive and costly. Automatic simplification is a field which attracts growing attention in Natural Language Processing, but, to the best of our knowledge, there are no simplification tools for Spanish. We present a prototype for automatic simplification, which shows that the most important structural simplification operations can be successfully treated with an approach based on rules which can potentially be improved by statistical methods. For the development of this prototype we carried out a corpus study which aims at identifying the operations a text simplification system needs to carry out in order to produce an output similar to what human editors produce when they simplify texts.
The relevance of syntactic dependency annotated corpora is nowadays unquestioned. However, a broad debate on the optimal set of dependency relation tags did not take place yet. As a result, largely varying tag sets of a largely varying size are used in different annotation initiatives. We propose a hierarchical dependency structure annotation schema that is more detailed and more flexible than the known annotation schemata. The schema allows us to choose the level of the desired detail of annotation, which facilitates the use of the schema for corpus annotation for different languages and for different NLP applications. Thanks to the inclusion of semantico-syntactic tags into the schema, we can annotate a corpus not only with syntactic dependency structures, but also with valency patterns as they are usually found in separate treebanks such as PropBank and NomBank. Semantico-syntactic tags and the level of detail of the schema furthermore facilitate the derivation of deep-syntactic and semantic annotations, leading to truly multilevel annotated dependency corpora. Such multilevel annotations can be readily used for the task of ML-based acquisition of grammar resources that map between the different levels of linguistic representation ― something which forms part of, for instance, any natural language text generator.
Hardly any other kind of text structures is as notoriously difficult to read as patents. This is first of all due to their abstract vocabulary and their very complex syntactic constructions. Especially the claims in a patent are a challenge: in accordance with international patent writing regulations, each claim must be rendered in a single sentence. As a result, sentences with more than 200 words are not uncommon. Therefore, paraphrasing of the claims in terms the user can understand is of high demand. We present a rule-based paraphrasing module that realizes paraphrasing of patent claims in English as a rewriting task. Prior to the rewriting proper, the module implies the stages of simplification and discourse and syntactic analyses. The rewriting makes use of a full-fledged text generator and consists in a number of genuine generation tasks such as aggregation, selection of referring expressions, choice of discourse markers and syntactic generation. As generator, we use the MATE-work bench, which is based on the Meaning-Text Theory of linguistics.