Large Language Models (LLMs) have showcased remarkable capabilities in various Natural Language Processing tasks. For automatic open-domain dialogue evaluation in particular, LLMs have been seamlessly integrated into evaluation frameworks, and together with human evaluation, compose the backbone of most evaluations. However, existing evaluation benchmarks often rely on outdated datasets and evaluate aspects like Fluency and Relevance, which fail to adequately capture the capabilities and limitations of state-of-the-art chatbot models. This paper critically examines current evaluation benchmarks, highlighting that the use of older response generators and quality aspects fail to accurately reflect modern chatbot capabilities. A small annotation experiment on a recent LLM-generated dataset (SODA) reveals that LLM evaluators such as GPT-4 struggle to detect actual deficiencies in dialogues generated by current LLM chatbots.

Although human evaluation remains the gold standard for open-domain dialogue evaluation, the growing popularity of automated evaluation using Large Language Models (LLMs) has also extended to dialogue. However, most frameworks leverage benchmarks that assess older chatbots on aspects such as fluency and relevance, which are not reflective of the challenges associated with contemporary models. In fact, a qualitative analysis on Soda. (Kim et al., 2023), a GPT-3.5 generated dialogue dataset, suggests that current chatbots may exhibit several recurring issues related to coherence and commonsense knowledge, but generally produce highly fluent and relevant responses.Noting the aforementioned limitations, this paper introduces Soda-Eval, an annotated dataset based on Soda that covers over 120K turn-level assessments across 10K dialogues, where the annotations were generated by GPT-4. Using Soda-Eval as a benchmark, we then study the performance of several open-access instruction-tuned LLMs, finding that dialogue evaluation remains challenging. Fine-tuning these models improves performance over few-shot inferences, both in terms of correlation and explanation.

Despite being heralded as the new standard for dialogue evaluation, the closed-source nature of GPT-4 poses challenges for the community. Motivated by the need for lightweight, open source, and multilingual dialogue evaluators, this paper introduces GenResCoh (Generated Responses targeting Coherence). GenResCoh is a novel LLM generated dataset comprising over 130k negative and positive responses and accompanying explanations seeded from XDailyDialog and XPersona covering English, French, German, Italian, and Chinese. Leveraging GenResCoh, we propose ECoh (Evaluation of Coherence), a family of evaluators trained to assess response coherence across multiple languages. Experimental results demonstrate that ECoh achieves multilingual detection capabilities superior to the teacher model (GPT-3.5-Turbo) on GenResCoh, despite being based on a much smaller architecture. Furthermore, the explanations provided by ECoh closely align in terms of quality with those generated by the teacher model.

The main limiting factor in the development of robust multilingual open-domain dialogue evaluation metrics is the lack of multilingual data and the limited availability of open-sourced multilingual dialogue systems. In this work, we propose a workaround for this lack of data by leveraging a strong multilingual pretrained encoder-based Language Model and augmenting existing English dialogue data using Machine Translation. We empirically show that the naive approach of finetuning a pretrained multilingual encoder model with translated data is insufficient to outperform the strong baseline of finetuning a multilingual model with only source data. Instead, the best approach consists in the careful curation of translated data using MT Quality Estimation metrics, excluding low quality translations that hinder its performance.

Despite significant research effort in the development of automatic dialogue evaluation metrics, little thought is given to evaluating dialogues other than in English. At the same time, ensuring metrics are invariant to semantically similar responses is also an overlooked topic. In order to achieve the desired properties of robustness and multilinguality for dialogue evaluation metrics, we propose a novel framework that takes advantage of the strengths of current evaluation models with the newly-established paradigm of prompting Large Language Models (LLMs). Empirical results show our framework achieves state of the art results in terms of mean Spearman correlation scores across several benchmarks and ranks first place on both the Robust and Multilingual tasks of the DSTC11 Track 4 “Automatic Evaluation Metrics for Open-Domain Dialogue Systems”, proving the evaluation capabilities of prompted LLMs.

Task-oriented conversational datasets often lack topic variability and linguistic diversity. However, with the advent of Large Language Models (LLMs) pretrained on extensive, multilingual and diverse text data, these limitations seem overcome. Nevertheless, their generalisability to different languages and domains in dialogue applications remains uncertain without benchmarking datasets. This paper presents a holistic annotation approach for emotion and conversational quality in the context of bilingual customer support conversations. By performing annotations that take into consideration the complete instances that compose a conversation, one can form a broader perspective of the dialogue as a whole. Furthermore, it provides a unique and valuable resource for the development of text classification models. To this end, we present benchmarks for Emotion Recognition and Dialogue Quality Estimation and show that further research is needed to leverage these models in a production setting.

Despite considerable advances in open-domain neural dialogue systems, their evaluation remains a bottleneck. Several automated metrics have been proposed to evaluate these systems, however, they mostly focus on a single notion of quality, or, when they do combine several sub-metrics, they are computationally expensive. This paper attempts to solve the latter: QualityAdapt leverages the Adapter framework for the task of Dialogue Quality Estimation. Using well defined semi-supervised tasks, we train adapters for different subqualities and score generated responses with AdapterFusion. This compositionality provides an easy to adapt metric to the task at hand that incorporates multiple subqualities. It also reduces computational costs as individual predictions of all subqualities are obtained in a single forward pass. This approach achieves comparable results to state-of-the-art metrics on several datasets, whilst keeping the previously mentioned advantages.

Crowdsourcing the collection of speech provides a scalable setting to access a customisable demographic according to each dataset’s needs. The correctness of speaker metadata is especially relevant for speaker-centred collections - ones that require the collection of a fixed amount of data per speaker. This paper identifies two different types of misalignment present in these collections: Multiple Accounts misalignment (different contributors map to the same speaker), and Multiple Speakers misalignment (multiple speakers map to the same contributor). Based on state-of-the-art approaches to Speaker Verification, this paper proposes an unsupervised method for measuring speaker metadata plausibility of a collection, i.e., evaluating the match (or lack thereof) between contributors and speakers. The solution presented is composed of an embedding extractor and a clustering module. Results indicate high precision in automatically classifying contributor alignment (>0.94).

The automation of the diagnosis and monitoring of speech affecting diseases in real life situations, such as Depression or Parkinson’s disease, depends on the existence of rich and large datasets that resemble real life conditions, such as those collected from in-the-wild multimedia repositories like YouTube. However, the cost of manually labeling these large datasets can be prohibitive. In this work, we propose to overcome this problem by automating the annotation process, without any requirements for human intervention. We formulate the annotation problem as a Multiple Instance Learning (MIL) problem, and propose a novel solution that is based on end-to-end differentiable neural networks. Our solution has the additional advantage of generalizing the MIL framework to more scenarios where the data is stil organized in bags but does not meet the MIL bag label conditions. We demonstrate the performance of the proposed method in labeling the in-the-Wild Speech Medical (WSM) Corpus, using simple textual cues extracted from videos and their metadata. Furthermore we show what is the contribution of each type of textual cues for the final model performance, as well as study the influence of the size of the bags of instances in determining the difficulty of the learning problem

This paper presents SPA, a web-based Speech Analytics platform that integrates several speech processing modules and that makes it possible to use them through the web. It was developed with the aim of facilitating the usage of the modules, without the need to know about software dependencies and specific configurations. Apart from being accessed by a web-browser, the platform also provides a REST API for easy integration with other applications. The platform is flexible, scalable, provides authentication for access restrictions, and was developed taking into consideration the time and effort of providing new services. The platform is still being improved, but it already integrates a considerable number of audio and text processing modules, including: Automatic transcription, speech disfluency classification, emotion detection, dialog act recognition, age and gender classification, non-nativeness detection, hyper-articulation detection, dialog act recognition, and two external modules for feature extraction and DTMF detection. This paper describes the SPA architecture, presents the already integrated modules, and provides a detailed description for the ones most recently integrated.

This paper presents a linguistic revision process of a speech corpus of Portuguese broadcast news focusing on metadata annotation for rich transcription, and reports on the impact of the new data on the performance for several modules. The main focus of the revision process consisted on annotating and revising structural metadata events, such as disfluencies and punctuation marks. The resultant revised data is now being extensively used, and was of extreme importance for improving the performance of several modules, especially the punctuation and capitalization modules, but also the speech recognition system, and all the subsequent modules. The resultant data has also been recently used in disfluency studies across domains.

This paper presents 3 sets of OpenLogos resources, namely the English-German, the English-French, and the English-Italian bilingual dictionaries. In addition to the usual information on part-of-speech, gender, and number for nouns, offered by most dictionaries currently available, OpenLogos bilingual dictionaries have some distinctive features that make them unique: they contain cross-language morphological information (inflectional and derivational), semantico-syntactic knowledge, indication of the head word in multiword units, information about whether a source word corresponds to an homograph, information about verb auxiliaries, alternate words (i.e., predicate or process nouns), causatives, reflexivity, verb aspect, among others. The focal point of the paper will be the semantico-syntactic knowledge that is important for disambiguation and translation precision. The resources are publicly available at the METANET platform for free use by the research community.

This paper presents a systematic human evaluation of translations of English support verb constructions produced by a rule-based machine translation (RBMT) system (OpenLogos) and a statistical machine translation (SMT) system (Google Translate) for five languages: French, German, Italian, Portuguese and Spanish. We classify support verb constructions by means of their syntactic structure and semantic behavior and present a qualitative analysis of their translation errors. The study aims to verify how machine translation (MT) systems translate fine-grained linguistic phenomena, and how well-equipped they are to produce high-quality translation. Another goal of the linguistically motivated quality analysis of SVC raw output is to reinforce the need for better system hybridization, which leverages the strengths of RBMT to the benefit of SMT, especially in improving the translation of multiword units. Taking multiword units into account, we propose an effective method to achieve MT hybridization based on the integration of semantico-syntactic knowledge into SMT.

In Statistical Machine Translation, words that were not seen during training are unknown words, that is, words that the system will not know how to translate. In this paper we contribute to this research problem by profiting from orthographic cues given by words. Thus, we report a study of the impact of word distance metrics in cognates' detection and, in addition, on the possibility of obtaining possible translations of unknown words through Logical Analogy. Our approach is tested in the translation of corpora from Portuguese to English (and vice-versa).

This work describes a process to extract Named Entity (NE) translations from the text available in web links (anchor texts). It translates a NE by retrieving a list of web documents in the target language, extracting the anchor texts from the links to those documents and finding the best translation from the anchor texts, using a combination of features, some of which, are specific to anchor texts. Experiments performed on a manually built corpora, suggest that over 70% of the NEs, ranging from unpopular to popular entities, can be translated correctly using sorely anchor texts. Tests on a Machine Translation task indicate that the system can be used to improve the quality of the translations of state-of-the-art statistical machine translation systems.

In this paper we describe the Instituto de Engenharia de Sistemas e Computadores Investigac ̧a ̃o e Desenvolvimento (INESC-ID) system that participated in the IWSLT 2010 evaluation campaign. Our main goal for this evaluation was to employ several state-of-the-art methods applied to phrase-based machine translation in order to improve the translation quality. Aside from the IBM M4 alignment model, two constrained alignment models were tested, which produced better overall results. These results were further improved by using weighted alignment matrixes during phrase extraction, rather than the single best alignment. Finally, we tested several filters that ruled out phrase pairs based on puntuation. Our system was evaluated on the BTEC and DIALOG tasks, having achieved a better overall ranking in the DIALOG task.

Phrase-based systems deeply depend on the quality of their phrase tables and therefore, the process of phrase extraction is always a fundamental step. In this paper we present a general and extensible phrase extraction algorithm, where we have highlighted several control points. The instantiation of these control points allows the simulation of previous approaches, as in each one of these points different strategies/heuristics can be tested. We show how previous approaches fit in this algorithm, compare several of them and, in addition, we propose alternative heuristics, showing their impact on the final translation results. Considering two different test scenarios from the IWSLT 2010 competition (BTEC, Fr-En and DIALOG, Cn-En), we have obtained an improvement in the results of 2.4 and 2.8 BLEU points, respectively.

This paper describes the corpus of university lectures that has been recorded in European Portuguese, and some of the recognition experiments we have done with it. The highly specific topic domain and the spontaneous speech nature of the lectures are two of the most challenging problems. Lexical and language model adaptation proved difficult given the scarcity of domain material in Portuguese, but improvements can be achieved with unsupervised acoustic model adaptation. From the point of view of the study of spontaneous speech characteristics, namely disflluencies, the LECTRA corpus has also proved a very valuable resource.