Fausto Giunchiglia

Khuyagbaatar Batsuren, Gábor Bella, Aryaman Arora et al. · eth-zurich, stanford

The SIGMORPHON 2022 shared task on morpheme segmentation challenged systems to decompose a word into a sequence of morphemes and covered most types of morphology: compounds, derivations, and inflections. Subtask 1, word-level morpheme segmentation, covered 5 million words in 9 languages (Czech, English, Spanish, Hungarian, French, Italian, Russian, Latin, Mongolian) and received 13 system submissions from 7 teams and the best system averaged 97.29% F1 score across all languages, ranging English (93.84%) to Latin (99.38%). Subtask 2, sentence-level morpheme segmentation, covered 18,735 sentences in 3 languages (Czech, English, Mongolian), received 10 system submissions from 3 teams, and the best systems outperformed all three state-of-the-art subword tokenization methods (BPE, ULM, Morfessor2) by 30.71% absolute. To facilitate error analysis and support any type of future studies, we released all system predictions, the evaluation script, and all gold standard datasets.

Yonghao Liu, Mengyu Li, Ximing Li et al.

Node classification is an essential problem in graph learning. However, many models typically obtain unsatisfactory performance when applied to few-shot scenarios. Some studies have attempted to combine meta-learning with graph neural networks to solve few-shot node classification on graphs. Despite their promising performance, some limitations remain. First, they employ the node encoding mechanism of homophilic graphs to learn node embeddings, even in heterophilic graphs. Second, existing models based on meta-learning ignore the interference of randomness in the learning process. Third, they are trained using only limited labeled nodes within the specific task, without explicitly utilizing numerous unlabeled nodes. Finally, they treat almost all sampled tasks equally without customizing them for their uniqueness. To address these issues, we propose a novel framework for few-shot node classification called Meta-GPS++. Specifically, we first adopt an efficient method to learn discriminative node representations on homophilic and heterophilic graphs. Then, we leverage a prototype-based approach to initialize parameters and contrastive learning for regularizing the distribution of node embeddings. Moreover, we apply self-training to extract valuable information from unlabeled nodes. Additionally, we adopt S$^2$ (scaling & shifting) transformation to learn transferable knowledge from diverse tasks. The results on real-world datasets show the superiority of Meta-GPS++. Our code is available here.

Andrea Bontempelli, Stefano Teso, Katya Tentori et al.

Part-prototype Networks (ProtoPNets) are concept-based classifiers designed to achieve the same performance as black-box models without compromising transparency. ProtoPNets compute predictions based on similarity to class-specific part-prototypes learned to recognize parts of training examples, making it easy to faithfully determine what examples are responsible for any target prediction and why. However, like other models, they are prone to picking up confounders and shortcuts from the data, thus suffering from compromised prediction accuracy and limited generalization. We propose ProtoPDebug, an effective concept-level debugger for ProtoPNets in which a human supervisor, guided by the model's explanations, supplies feedback in the form of what part-prototypes must be forgotten or kept, and the model is fine-tuned to align with this supervision. Our experimental evaluation shows that ProtoPDebug outperforms state-of-the-art debuggers for a fraction of the annotation cost. An online experiment with laypeople confirms the simplicity of the feedback requested to the users and the effectiveness of the collected feedback for learning confounder-free part-prototypes. ProtoPDebug is a promising tool for trustworthy interactive learning in critical applications, as suggested by a preliminary evaluation on a medical decision making task.

Paula Helm, Gábor Bella, Gertraud Koch et al.

It is well known that AI-based language technology -- large language models, machine translation systems, multilingual dictionaries, and corpora -- is currently limited to 2 to 3 percent of the world's most widely spoken and/or financially and politically best supported languages. In response, recent research efforts have sought to extend the reach of AI technology to ``underserved languages.'' In this paper, we show that many of these attempts produce flawed solutions that adhere to a hard-wired representational preference for certain languages, which we call techno-linguistic bias. Techno-linguistic bias is distinct from the well-established phenomenon of linguistic bias as it does not concern the languages represented but rather the design of the technologies. As we show through the paper, techno-linguistic bias can result in systems that can only express concepts that are part of the language and culture of dominant powers, unable to correctly represent concepts from other communities. We argue that at the root of this problem lies a systematic tendency of technology developer communities to apply a simplistic understanding of diversity which does not do justice to the more profound differences that languages, and ultimately the communities that speak them, embody. Drawing on the concept of epistemic injustice, we point to the broader sociopolitical consequences of the bias we identify and show how it can lead not only to a disregard for valuable aspects of diversity but also to an under-representation of the needs and diverse worldviews of marginalized language communities.

Gábor Bella, Erdenebileg Byambadorj, Yamini Chandrashekar et al.

The Universal Knowledge Core (UKC) is a large multilingual lexical database with a focus on language diversity and covering over a thousand languages. The aim of the database, as well as its tools and data catalogue, is to make the somewhat abstract notion of diversity visually understandable for humans and formally exploitable by machines. The UKC website lets users explore millions of individual words and their meanings, but also phenomena of cross-lingual convergence and divergence, such as shared interlingual meanings, lexicon similarities, cognate clusters, or lexical gaps. The UKC LiveLanguage Catalogue, in turn, provides access to the underlying lexical data in a computer-processable form, ready to be reused in cross-lingual applications.

Hadi Khalilia, Gábor Bella, Abed Alhakim Freihat et al.

Languages are known to describe the world in diverse ways. Across lexicons, diversity is pervasive, appearing through phenomena such as lexical gaps and untranslatability. However, in computational resources, such as multilingual lexical databases, diversity is hardly ever represented. In this paper, we introduce a method to enrich computational lexicons with content relating to linguistic diversity. The method is verified through two large-scale case studies on kinship terminology, a domain known to be diverse across languages and cultures: one case study deals with seven Arabic dialects, while the other one with three Indonesian languages. Our results, made available as browseable and downloadable computational resources, extend prior linguistics research on kinship terminology, and provide insight into the extent of diversity even within linguistically and culturally close communities.

Fausto Giunchiglia, Gabor Bella, Nandu Chandran Nair et al.

In today's multilingual lexical databases, the majority of the world's languages are under-represented. Beyond a mere issue of resource incompleteness, we show that existing lexical databases have structural limitations that result in a reduced expressivity on culturally-specific words and in mapping them across languages. In particular, the lexical meaning space of dominant languages, such as English, is represented more accurately while linguistically or culturally diverse languages are mapped in an approximate manner. Our paper assesses state-of-the-art multilingual lexical databases and evaluates their strengths and limitations with respect to their expressivity on lexical phenomena of linguistic diversity.

Temuulen Khishigsuren, Gábor Bella, Khuyagbaatar Batsuren et al.

This paper describes a method to enrich lexical resources with content relating to linguistic diversity, based on knowledge from the field of lexical typology. We capture the phenomenon of diversity through the notions of lexical gap and language-specific word and use a systematic method to infer gaps semi-automatically on a large scale. As a first result obtained for the domain of kinship terminology, known to be very diverse throughout the world, we publish a lexico-semantic resource consisting of 198 domain concepts, 1,911 words, and 37,370 gaps covering 699 languages. We see potential in the use of resources such as ours for the improvement of a variety of cross-lingual NLP tasks, which we demonstrate through a downstream application for the evaluation of machine translation systems.

Fausto Giunchiglia, Mayukh Bagchi, Xiaolei Diao

We discuss two kinds of semantics relevant to Computer Vision (CV) systems - Visual Semantics and Lexical Semantics. While visual semantics focus on how humans build concepts when using vision to perceive a target reality, lexical semantics focus on how humans build concepts of the same target reality through the use of language. The lack of coincidence between visual and lexical semantics, in turn, has a major impact on CV systems in the form of the Semantic Gap Problem (SGP). The paper, while extensively exemplifying the lack of coincidence as above, introduces a general, domain-agnostic methodology to enforce alignment between visual and lexical semantics.

Fausto Giunchiglia, Mayukh Bagchi

Semantic Heterogeneity is conventionally understood as the existence of variance in the representation of a target reality when modelled, by independent parties, in different databases, schemas and/ or data. We argue that the mere encoding of variance, while being necessary, is not sufficient enough to deal with the problem of representational heterogeneity, given that it is also necessary to encode the unifying basis on which such variance is manifested. To that end, this paper introduces a notion of Representation Heterogeneity in terms of the co-occurrent notions of Representation Unity and Representation Diversity. We have representation unity when two heterogeneous representations model the same target reality, representation diversity otherwise. In turn, this paper also highlights how these two notions get instantiated across the two layers of any representation, i.e., Language and Knowledge.

Andrea Bontempelli, Marcelo Rodas Britez, Xiaoyue Li et al.

We focus on the development of AIs which live in lifelong symbiosis with a human. The key prerequisite for this task is that the AI understands - at any moment in time - the personal situational context that the human is in. We outline the key challenges that this task brings forth, namely (i) handling the human-like and ego-centric nature of the the user's context, necessary for understanding and providing useful suggestions, (ii) performing lifelong context recognition using machine learning in a way that is robust to change, and (iii) maintaining alignment between the AI's and human's representations of the world through continual bidirectional interaction. In this short paper, we summarize our recent attempts at tackling these challenges, discuss the lessons learned, and highlight directions of future research. The main take-away message is that pursuing this project requires research which lies at the intersection of knowledge representation and machine learning. Neither technology can achieve this goal without the other.

Fausto Giunchiglia, Mayukh Bagchi, Xiaolei Diao

Recent work in Machine Learning and Computer Vision has highlighted the presence of various types of systematic flaws inside ground truth object recognition benchmark datasets. Our basic tenet is that these flaws are rooted in the many-to-many mappings which exist between the visual information encoded in images and the intended semantics of the labels annotating them. The net consequence is that the current annotation process is largely under-specified, thus leaving too much freedom to the subjective judgment of annotators. In this paper, we propose vTelos, an integrated Natural Language Processing, Knowledge Representation, and Computer Vision methodology whose main goal is to make explicit the (otherwise implicit) intended annotation semantics, thus minimizing the number and role of subjective choices. A key element of vTelos is the exploitation of the WordNet lexico-semantic hierarchy as the main means for providing the meaning of natural language labels and, as a consequence, for driving the annotation of images based on the objects and the visual properties they depict. The methodology is validated on images populating a subset of the ImageNet hierarchy.

Gábor Bella, Paula Helm, Gertraud Koch et al.

It is a well-known fact that current AI-based language technology -- language models, machine translation systems, multilingual dictionaries and corpora -- focuses on the world's 2-3% most widely spoken languages. Recent research efforts have attempted to expand the coverage of AI technology to `under-resourced languages.' The goal of our paper is to bring attention to a phenomenon that we call linguistic bias: multilingual language processing systems often exhibit a hardwired, yet usually involuntary and hidden representational preference towards certain languages. Linguistic bias is manifested in uneven per-language performance even in the case of similar test conditions. We show that biased technology is often the result of research and development methodologies that do not do justice to the complexity of the languages being represented, and that can even become ethically problematic as they disregard valuable aspects of diversity as well as the needs of the language communities themselves. As our attempt at building diversity-aware language resources, we present a new initiative that aims at reducing linguistic bias through both technological design and methodology, based on an eye-level collaboration with local communities.

Fausto Giunchiglia, Simone Bocca, Mattia Fumagalli et al.

More and more, with the growing focus on large scale analytics, we are confronted with the need of integrating data from multiple sources. The problem is that these data are impossible to reuse as-is. The net result is high cost, with the further drawback that the resulting integrated data will again be hardly reusable as-is. iTelos is a general purpose methodology aiming at minimizing the effects of this process. The intuition is that data will be treated differently based on their popularity: the more a certain set of data have been reused, the more they will be reused and the less they will be changed across reuses, thus decreasing the overall data preprocessing costs, while increasing backward compatibility and future sharing

Mattia Fumagalli, Marco Boffo, Daqian Shi et al.

One of the significant barriers to the training of statistical models on knowledge graphs is the difficulty that scientists have in finding the best input data to address their prediction goal. In addition to this, a key challenge is to determine how to manipulate these relational data, which are often in the form of particular triples (i.e., subject, predicate, object), to enable the learning process. Currently, many high-quality catalogs of knowledge graphs, are available. However, their primary goal is the re-usability of these resources, and their interconnection, in the context of the Semantic Web. This paper describes the LiveSchema initiative, namely, a first version of a gateway that has the main scope of leveraging the gold mine of data collected by many existing catalogs collecting relational data like ontologies and knowledge graphs. At the current state, LiveSchema contains - 1000 datasets from 4 main sources and offers some key facilities, which allow to: i) evolving LiveSchema, by aggregating other source catalogs and repositories as input sources; ii) querying all the collected resources; iii) transforming each given dataset into formal concept analysis matrices that enable analysis and visualization services; iv) generating models and tensors from each given dataset.

Fausto Giunchiglia, Xiaolei Diao, Mayukh Bagchi

Data quality is critical for multimedia tasks, while various types of systematic flaws are found in image benchmark datasets, as discussed in recent work. In particular, the existence of the semantic gap problem leads to a many-to-many mapping between the information extracted from an image and its linguistic description. This unavoidable bias further leads to poor performance on current computer vision tasks. To address this issue, we introduce a Knowledge Representation (KR)-based methodology to provide guidelines driving the labeling process, thereby indirectly introducing intended semantics in ML models. Specifically, an iterative refinement-based annotation method is proposed to optimize data labeling by organizing objects in a classification hierarchy according to their visual properties, ensuring that they are aligned with their linguistic descriptions. Preliminary results verify the effectiveness of the proposed method.

Daqian Shi, Xiaoyue Li, Fausto Giunchiglia

A common solution to the semantic heterogeneity problem is to perform knowledge graph (KG) extension exploiting the information encoded in one or more candidate KGs, where the alignment between the reference KG and candidate KGs is considered the critical procedure. However, existing KG alignment methods mainly rely on entity type (etype) label matching as a prerequisite, which is poorly performing in practice or not applicable in some cases. In this paper, we design a machine learning-based framework for KG extension, including an alternative novel property-based alignment approach that allows aligning etypes on the basis of the properties used to define them. The main intuition is that it is properties that intentionally define the etype, and this definition is independent of the specific label used to name an etype, and of the specific hierarchical schema of KGs. Compared with the state-of-the-art, the experimental results show the validity of the KG alignment approach and the superiority of the proposed KG extension framework, both quantitatively and qualitatively.

Mattia Fumagalli, Marco Boffo, Daqian Shi et al.

One of the major barriers to the training of algorithms on knowledge graph schemas, such as vocabularies or ontologies, is the difficulty that scientists have in finding the best input resource to address the target prediction tasks. In addition to this, a key challenge is to determine how to manipulate (and embed) these data, which are often in the form of particular triples (i.e., subject, predicate, object), to enable the learning process. In this paper, we describe the LiveSchema initiative, namely a gateway that offers a family of services to easily access, analyze, transform and exploit knowledge graph schemas, with the main goal of facilitating the reuse of these resources in machine learning use cases. As an early implementation of the initiative, we also advance an online catalog, which relies on more than 800 resources, with the first set of example services.

Xiaosong Han, Ke Chen, Xindi Dai et al.

In real-world deployment, LLMs are often adapted continually across tasks to keep LLMs up-to-date in production, where new fine-tuning should preserve previously learned skills. However, indiscriminately mixing tasks can dilute task specialization, while sequential fine-tuning (full-parameter or low rank adaptation) often causes catastrophic forgetting due to destructive overwriting. Replay-based continual tuning and maintaining separate task-specific adapters can mitigate forgetting, but introduce additional compute, storage, and management overhead. Recognizing the redundancy of LLM parameters for any single task, we reframe continual task adaptation as task-specific parameter discovery via adaptation-aware probing: a short warm-start probe exposes a task's adaptation trace, enabling us to identify and isolate the small subset of parameters essential for each task to mitigate catastrophic forgetting. Building on this view, we introduce TRACE, a novel approach for discovering Task-specific paRameters via Adaptation-aware probing for Continual finE-tuning. We perform a short warm-start fine-tune to derive task-specific core parameters by comparing the warm-started and pre-trained models. Core parameters are identified via two strategies: importance scoring (L$_2$ norm and Fisher Information) and specificity analysis (cosine similarity of parameter updates). In continual fine-tuning settings, only the active task's core parameters are updated while others remain frozen, preserving prior knowledge. We conduct extensive experiments across multiple standard benchmarks to demonstrate the superior performance of our proposed method. Additionally, we validate the generalization of our method through a cross-model and scale transferability study, demonstrating a "small-to-large" paradigm that guides the fine-tuning of large-scale models under resource constraints.

Rui Song, Fausto Giunchiglia, Yingji Li et al.

Despite large-scale pre-trained language models have achieved striking results for text classificaion, recent work has raised concerns about the challenge of shortcut learning. In general, a keyword is regarded as a shortcut if it creates a superficial association with the label, resulting in a false prediction. Conversely, shortcut learning can be mitigated if the model relies on robust causal features that help produce sound predictions. To this end, many studies have explored post-hoc interpretable methods to mine shortcuts and causal features for robustness and generalization. However, most existing methods focus only on single word in a sentence and lack consideration of word-group, leading to wrong causal features. To solve this problem, we propose a new Word-Group mining approach, which captures the causal effect of any keyword combination and orders the combinations that most affect the prediction. Our approach bases on effective post-hoc analysis and beam search, which ensures the mining effect and reduces the complexity. Then, we build a counterfactual augmentation method based on the multiple word-groups, and use an adaptive voting mechanism to learn the influence of different augmentated samples on the prediction results, so as to force the model to pay attention to effective causal features. We demonstrate the effectiveness of the proposed method by several tasks on 8 affective review datasets and 4 toxic language datasets, including cross-domain text classificaion, text attack and gender fairness test.

Daqian Shi, Fausto Giunchiglia

The mainstream approach to the development of ontologies is merging ontologies encoding different information, where one of the major difficulties is that the heterogeneity motivates the ontology merging but also limits high-quality merging performance. Thus, the entity type (etype) recognition task is proposed to deal with such heterogeneity, aiming to infer the class of entities and etypes by exploiting the information encoded in ontologies. In this paper, we introduce a property-based approach that allows recognizing etypes on the basis of the properties used to define them. From an epistemological point of view, it is in fact properties that characterize entities and etypes, and this definition is independent of the specific labels and hierarchical schemas used to define them. The main contribution consists of a set of property-based metrics for measuring the contextual similarity between etypes and entities, and a machine learning-based etype recognition algorithm exploiting the proposed similarity metrics. Compared with the state-of-the-art, the experimental results show the validity of the similarity metrics and the superiority of the proposed etype recognition algorithm.

Mattia Fumagalli, Daqian Shi, Fausto Giunchiglia

The main goal of this paper is to evaluate knowledge base schemas, modeled as a set of entity types, each such type being associated with a set of properties, according to their focus. We intuitively model the notion of focus as ''the state or quality of being relevant in storing and retrieving information''. This definition of focus is adapted from the notion of ''categorization purpose'', as first defined in cognitive psychology, thus giving us a high level of understandability on the side of users. In turn, this notion is formalized based on a set of knowledge metrics that, for any given focus, rank knowledge base schemas according to their quality. We apply the proposed methodology to more than 200 state-of-the-art knowledge base schemas. The experimental results show the utility of our approach

Yonghao Liu, Fausto Giunchiglia, Ximing Li et al.

Graph few-shot learning has garnered significant attention for its ability to rapidly adapt to downstream tasks with limited labeled data, sparking considerable interest among researchers. Recent advancements in graph few-shot learning models have exhibited superior performance across diverse applications. Despite their successes, several limitations still exist. First, existing models in the meta-training phase predominantly focus on instance-level features within tasks, neglecting crucial set-level features essential for distinguishing between different categories. Second, these models often utilize query sets directly on classifiers trained with support sets containing only a few labeled examples, overlooking potential distribution shifts between these sets and leading to suboptimal performance. Finally, previous models typically require necessitate abundant labeled data from base classes to extract transferable knowledge, which is typically infeasible in real-world scenarios. To address these issues, we propose a novel model named STAR, which leverages Set funcTions and optimAl tRansport for enhancing unsupervised graph few-shot learning. Specifically, STAR utilizes expressive set functions to obtain set-level features in an unsupervised manner and employs optimal transport principles to align the distributions of support and query sets, thereby mitigating distribution shift effects. Theoretical analysis demonstrates that STAR can capture more task-relevant information and enhance generalization capabilities. Empirically, extensive experiments across multiple datasets validate the effectiveness of STAR. Our code can be found here.

Yonghao Liu, Yajun Wang, Chunli Guo et al.

Graph few-shot learning has attracted increasing attention due to its ability to rapidly adapt models to new tasks with only limited labeled nodes. Despite the remarkable progress made by existing graph few-shot learning methods, several key limitations remain. First, most current approaches rely on predefined and unified graph filters (e.g., low-pass or high-pass filters) to globally enhance or suppress node frequency signals. Such fixed spectral operations fail to account for the heterogeneity of local topological structures inherent in real-world graphs. Moreover, these methods often assume that the support and query sets are drawn from the same distribution. However, under few-shot conditions, the limited labeled data in the support set may not sufficiently capture the complex distribution of the query set, leading to suboptimal generalization. To address these challenges, we propose GRACE, a novel Graph few-shot leaRning framework that integrates Adaptive spectrum experts with Cross-sEt distribution calibration techniques. Theoretically, the proposed approach enhances model generalization by adapting to both local structural variations and cross-set distribution calibration. Empirically, GRACE consistently outperforms state-of-the-art baselines across a wide range of experimental settings. Our code can be found here.

Yonghao Liu, Mengyu Li, Di Liang et al.

Natural Language Inference (NLI) is a crucial task in natural language processing that involves determining the relationship between two sentences, typically referred to as the premise and the hypothesis. However, traditional NLI models solely rely on the semantic information inherent in independent sentences and lack relevant situational visual information, which can hinder a complete understanding of the intended meaning of the sentences due to the ambiguity and vagueness of language. To address this challenge, we propose an innovative ScenaFuse adapter that simultaneously integrates large-scale pre-trained linguistic knowledge and relevant visual information for NLI tasks. Specifically, we first design an image-sentence interaction module to incorporate visuals into the attention mechanism of the pre-trained model, allowing the two modalities to interact comprehensively. Furthermore, we introduce an image-sentence fusion module that can adaptively integrate visual information from images and semantic information from sentences. By incorporating relevant visual information and leveraging linguistic knowledge, our approach bridges the gap between language and vision, leading to improved understanding and inference capabilities in NLI tasks. Extensive benchmark experiments demonstrate that our proposed ScenaFuse, a scenario-guided approach, consistently boosts NLI performance.

Maria Kasinidou, Styliani Kleanthous, Matteo Busso et al.

With the surge in data-centric AI and its increasing capabilities, AI applications have become a part of our everyday lives. However, misunderstandings regarding their capabilities, limitations, and associated advantages and disadvantages are widespread. Consequently, in the university setting, there is a crucial need to educate not only computer science majors but also students from various disciplines about AI. In this experience report, we present an overview of an introductory course that we offered to students coming from different majors. Moreover, we discuss the assignments and quizzes of the course, which provided students with a firsthand experience of AI processes and insights into their learning patterns. Additionally, we provide a summary of the course evaluation, as well as students' performance. Finally, we present insights gained from teaching this course and elaborate on our future plans.

Rui Song, Fausto Giunchiglia, Yingji Li et al.

Cross-domain text classification aims to transfer models from label-rich source domains to label-poor target domains, giving it a wide range of practical applications. Many approaches promote cross-domain generalization by capturing domain-invariant features. However, these methods rely on unlabeled samples provided by the target domains, which renders the model ineffective when the target domain is agnostic. Furthermore, the models are easily disturbed by shortcut learning in the source domain, which also hinders the improvement of domain generalization ability. To solve the aforementioned issues, this paper proposes TACIT, a target domain agnostic feature disentanglement framework which adaptively decouples robust and unrobust features by Variational Auto-Encoders. Additionally, to encourage the separation of unrobust features from robust features, we design a feature distillation task that compels unrobust features to approximate the output of the teacher. The teacher model is trained with a few easy samples that are easy to carry potential unknown shortcuts. Experimental results verify that our framework achieves comparable results to state-of-the-art baselines while utilizing only source domain data.

Mengyu Li, Yonghao Liu, Fausto Giunchiglia et al.

Text classification is a crucial and fundamental task in natural language processing. Compared with the previous learning paradigm of pre-training and fine-tuning by cross entropy loss, the recently proposed supervised contrastive learning approach has received tremendous attention due to its powerful feature learning capability and robustness. Although several studies have incorporated this technique for text classification, some limitations remain. First, many text datasets are imbalanced, and the learning mechanism of supervised contrastive learning is sensitive to data imbalance, which may harm the model performance. Moreover, these models leverage separate classification branch with cross entropy and supervised contrastive learning branch without explicit mutual guidance. To this end, we propose a novel model named SharpReCL for imbalanced text classification tasks. First, we obtain the prototype vector of each class in the balanced classification branch to act as a representation of each class. Then, by further explicitly leveraging the prototype vectors, we construct a proper and sufficient target sample set with the same size for each class to perform the supervised contrastive learning procedure. The empirical results show the effectiveness of our model, which even outperforms popular large language models across several datasets.

Yonghao Liu, Mengyu Li, Wei Pang et al.

Short text classification, as a research subtopic in natural language processing, is more challenging due to its semantic sparsity and insufficient labeled samples in practical scenarios. We propose a novel model named MI-DELIGHT for short text classification in this work. Specifically, it first performs multi-source information (i.e., statistical information, linguistic information, and factual information) exploration to alleviate the sparsity issues. Then, the graph learning approach is adopted to learn the representation of short texts, which are presented in graph forms. Moreover, we introduce a dual-level (i.e., instance-level and cluster-level) contrastive learning auxiliary task to effectively capture different-grained contrastive information within massive unlabeled data. Meanwhile, previous models merely perform the main task and auxiliary tasks in parallel, without considering the relationship among tasks. Therefore, we introduce a hierarchical architecture to explicitly model the correlations between tasks. We conduct extensive experiments across various benchmark datasets, demonstrating that MI-DELIGHT significantly surpasses previous competitive models. It even outperforms popular large language models on several datasets.

Gertraud Koch, Gábor Bella, Paula Helm et al.

Language technology has the potential to facilitate intercultural communication through meaningful translations. However, the current state of language technology is deeply entangled with colonial knowledge due to path dependencies and neo-colonial tendencies in the global governance of artificial intelligence (AI). Language technology is a complex and emerging field that presents challenges for co-design interventions due to enfolding in assemblages of global scale and diverse sites and its knowledge intensity. This paper uses LiveLanguage, a lexical database, a set of services with particular emphasis on modelling language diversity and integrating small and minority languages, as an example to discuss and close the gap from pluriversal design theory to practice. By diversifying the concept of emerging technology, we can better approach language technology in global contexts. The paper presents a model comprising of five layers of technological activity. Each layer consists of specific practices and stakeholders, thus provides distinctive spaces for co-design interventions as mode of inquiry for de-linking, re-thinking and re-building language technology towards pluriversality. In that way, the paper contributes to reflecting the position of co-design in decolonising emergent technologies, and to integrating complex theoretical knowledge towards decoloniality into language technology design.

Yonghao Liu, Fausto Giunchiglia, Lan Huang et al.

Short text classification has gained significant attention in the information age due to its prevalence and real-world applications. Recent advancements in graph learning combined with contrastive learning have shown promising results in addressing the challenges of semantic sparsity and limited labeled data in short text classification. However, existing models have certain limitations. They rely on explicit data augmentation techniques to generate contrastive views, resulting in semantic corruption and noise. Additionally, these models only focus on learning the intrinsic consistency between the generated views, neglecting valuable discriminative information from other potential views. To address these issues, we propose a Simple graph contrastive learning framework for Short Text Classification (SimSTC). Our approach involves performing graph learning on multiple text-related component graphs to obtain multi-view text embeddings. Subsequently, we directly apply contrastive learning on these embeddings. Notably, our method eliminates the need for data augmentation operations to generate contrastive views while still leveraging the benefits of multi-view contrastive learning. Despite its simplicity, our model achieves outstanding performance, surpassing large language models on various datasets.

Rui Song, Yingji Li, Lida Shi et al.

Shortcut learning refers to the phenomenon where models employ simple, non-robust decision rules in practical tasks, which hinders their generalization and robustness. With the rapid development of large language models (LLMs) in recent years, an increasing number of studies have shown the impact of shortcut learning on LLMs. This paper provides a novel perspective to review relevant research on shortcut learning in In-Context Learning (ICL). It conducts a detailed exploration of the types of shortcuts in ICL tasks, their causes, available benchmarks, and strategies for mitigating shortcuts. Based on corresponding observations, it summarizes the unresolved issues in existing research and attempts to outline the future research landscape of shortcut learning.

Fausto Giunchiglia, Mayukh Bagchi

Knowledge Representation (KR) and facet-analytical Knowledge Organization (KO) have been the two most prominent methodologies of data and knowledge modelling in the Artificial Intelligence community and the Information Science community, respectively. KR boasts of a robust and scalable ecosystem of technologies to support knowledge modelling while, often, underemphasizing the quality of its models (and model-based data). KO, on the other hand, is less technology-driven but has developed a robust framework of guiding principles (canons) for ensuring modelling (and model-based data) quality. This paper elucidates both the KR and facet-analytical KO methodologies in detail and provides a functional mapping between them. Out of the mapping, the paper proposes an integrated KO-enriched KR methodology with all the standard components of a KR methodology plus the guiding canons of modelling quality provided by KO. The practical benefits of the methodological integration has been exemplified through a prominent case study of KR-based image annotation exercise.

Hadi Khalilia, Jahna Otterbacher, Gabor Bella et al.

Lexical-semantic resources (LSRs), such as online lexicons and wordnets, are fundamental to natural language processing applications as well as to fields such as linguistic anthropology and language preservation. In many languages, however, such resources suffer from quality issues: incorrect entries, incompleteness, but also the rarely addressed issue of bias towards the English language and Anglo-Saxon culture. Such bias manifests itself in the absence of concepts specific to the language or culture at hand, the presence of foreign (Anglo-Saxon) concepts, as well as in the lack of an explicit indication of untranslatability, also known as cross-lingual lexical gaps, when a term has no equivalent in another language. This paper proposes a novel crowdsourcing methodology for reducing bias in LSRs. Crowd workers compare lexemes from two languages, focusing on domains rich in lexical diversity, such as kinship or food. Our LingoGap crowdsourcing platform facilitates comparisons through microtasks identifying equivalent terms, language-specific terms, and lexical gaps across languages. We validated our method by applying it to two case studies focused on food-related terminology: (1) English and Arabic, and (2) Standard Indonesian and Banjarese. These experiments identified 2,140 lexical gaps in the first case study and 951 in the second. The success of these experiments confirmed the usability of our method and tool for future large-scale lexicon enrichment tasks.

Abed Alhakim Freihat, Hadi Khalilia, Gábor Bella et al.

High-quality WordNets are crucial for achieving high-quality results in NLP applications that rely on such resources. However, the wordnets of most languages suffer from serious issues of correctness and completeness with respect to the words and word meanings they define, such as incorrect lemmas, missing glosses and example sentences, or an inadequate, Western-centric representation of the morphology and the semantics of the language. Previous efforts have largely focused on increasing lexical coverage while ignoring other qualitative aspects. In this paper, we focus on the Arabic language and introduce a major revision of the Arabic WordNet that addresses multiple dimensions of lexico-semantic resource quality. As a result, we updated more than 58% of the synsets of the existing Arabic WordNet by adding missing information and correcting errors. In order to address issues of language diversity and untranslatability, we also extended the wordnet structure by new elements: phrasets and lexical gaps.

Manisha Mehta, Fausto Giunchiglia

This research offers a unique evaluation of how AI systems interpret the digital language of Generation Alpha (Gen Alpha, born 2010-2024). As the first cohort raised alongside AI, Gen Alpha faces new forms of online risk due to immersive digital engagement and a growing mismatch between their evolving communication and existing safety tools. Their distinct language, shaped by gaming, memes, and AI-driven trends, often conceals harmful interactions from both human moderators and automated systems. We assess four leading AI models (GPT-4, Claude, Gemini, and Llama 3) on their ability to detect masked harassment and manipulation within Gen Alpha discourse. Using a dataset of 100 recent expressions from gaming platforms, social media, and video content, the study reveals critical comprehension failures with direct implications for online safety. This work contributes: (1) a first-of-its-kind dataset capturing Gen Alpha expressions; (2) a framework to improve AI moderation systems for youth protection; (3) a multi-perspective evaluation including AI systems, human moderators, and parents, with direct input from Gen Alpha co-researchers; and (4) an analysis of how linguistic divergence increases youth vulnerability. Findings highlight the urgent need to redesign safety systems attuned to youth communication, especially given Gen Alpha reluctance to seek help when adults fail to understand their digital world. This study combines the insight of a Gen Alpha researcher with systematic academic analysis to address critical digital safety challenges.

Fausto Giunchiglia, Mayukh Bagchi, Subhashis Das

Knowledge Organization (KO) and Knowledge Representation (KR) have been the two mainstream methodologies of knowledge modelling in the Information Science community and the Artificial Intelligence community, respectively. The facet-analytical tradition of KO has developed an exhaustive set of guiding canons for ensuring quality in organising and managing knowledge but has remained limited in terms of technology-driven activities to expand its scope and services beyond the bibliographic universe of knowledge. KR, on the other hand, boasts of a robust ecosystem of technologies and technology-driven service design which can be tailored to model any entity or scale to any service in the entire universe of knowledge. This paper elucidates both the facet-analytical KO and KR methodologies in detail and provides a functional mapping between them. Out of the mapping, the paper proposes an integrated KR-enriched KO methodology with all the standard components of a KO methodology plus the advanced technologies provided by the KR approach. The practical benefits of the methodological integration has been exemplified through the flagship application of the Digital University at the University of Trento, Italy.

Simone Bocca, Alessio Zamboni, Gabor Bella et al.

When building a new application we are increasingly confronted with the need of reusing and integrating pre-existing knowledge. Nevertheless, it is a fact that this prior knowledge is virtually impossible to reuse as-is. This is true also in domains, e.g., eHealth, where a lot of effort has been put into developing high-quality standards and reference ontologies, e.g. FHIR1. In this paper, we propose an integrated methodology, called iTelos, which enables data and knowledge reuse towards the construction of Interoperable Electronic Health Records (iEHR). The key intuition is that the data level and the schema level of an application should be developed independently, thus allowing for maximum flexibility in the reuse of the prior knowledge, but under the overall guidance of the needs to be satisfied, formalized as competence queries. This intuition is implemented by codifying all the requirements, including those concerning reuse, as part of a purpose defined a priori, which is then used to drive a middle-out development process where the application schema and data are continuously aligned. The proposed methodology is validated through its application to a large-scale case study.

Luca Erculiani, Andrea Bontempelli, Andrea Passerini et al.

We are interested in aligning how people think about objects and what machines perceive, meaning by this the fact that object recognition, as performed by a machine, should follow a process which resembles that followed by humans when thinking of an object associated with a certain concept. The ultimate goal is to build systems which can meaningfully interact with their users, describing what they perceive in the users' own terms. As from the field of Lexical Semantics, humans organize the meaning of words in hierarchies where the meaning of, e.g., a noun, is defined in terms of the meaning of a more general noun, its genus, and of one or more differentiating properties, its differentia. The main tenet of this paper is that object recognition should implement a hierarchical process which follows the hierarchical semantic structure used to define the meaning of words. We achieve this goal by implementing an algorithm which, for any object, recursively recognizes its visual genus and its visual differentia. In other words, the recognition of an object is decomposed in a sequence of steps where the locally relevant visual features are recognized. This paper presents the algorithm and a first evaluation.

Laura Schelenz, Ivano Bison, Matteo Busso et al.

Diversity-aware platform design is a paradigm that responds to the ethical challenges of existing social media platforms. Available platforms have been criticized for minimizing users' autonomy, marginalizing minorities, and exploiting users' data for profit maximization. This paper presents a design solution that centers the well-being of users. It presents the theory and practice of designing a diversity-aware platform for social relations. In this approach, the diversity of users is leveraged in a way that allows like-minded individuals to pursue similar interests or diverse individuals to complement each other in a complex activity. The end users of the envisioned platform are students, who participate in the design process. Diversity-aware platform design involves numerous steps, of which two are highlighted in this paper: 1) defining a framework and operationalizing the "diversity" of students, 2) collecting "diversity" data to build diversity-aware algorithms. The paper further reflects on the ethical challenges encountered during the design of a diversity-aware platform.

Fausto Giunchiglia, Mayukh Bagchi, Xiaolei Diao

Recent work in Machine Learning and Computer Vision has provided evidence of systematic design flaws in the development of major object recognition benchmark datasets. One such example is ImageNet, wherein, for several categories of images, there are incongruences between the objects they represent and the labels used to annotate them. The consequences of this problem are major, in particular considering the large number of machine learning applications, not least those based on Deep Neural Networks, that have been trained on these datasets. In this paper we posit the problem to be the lack of a knowledge representation (KR) methodology providing the foundations for the construction of these ground truth benchmark datasets. Accordingly, we propose a solution articulated in three main steps: (i) deconstructing the object recognition process in four ordered stages grounded in the philosophical theory of teleosemantics; (ii) based on such stratification, proposing a novel four-phased methodology for organizing objects in classification hierarchies according to their visual properties; and (iii) performing such classification according to the faceted classification paradigm. The key novelty of our approach lies in the fact that we construct the classification hierarchies from visual properties exploiting visual genus-differentiae, and not from linguistically grounded properties. The proposed approach is validated by a set of experiments on the ImageNet hierarchy of musical experiments.

Fausto Giunchiglia, Mayukh Bagchi

We base our work on the teleosemantic modelling of concepts as abilities implementing the distinct functions of recognition and classification. Accordingly, we model two types of concepts - substance concepts suited for object recognition exploiting visual properties, and classification concepts suited for classification of substance concepts exploiting linguistically grounded properties. The goal in this paper is to demonstrate that object recognition can be construed as classification via visual properties, as distinct from work in mainstream computer vision. Towards that, we present an object recognition process based on Ranganathan's four-phased faceted knowledge organization process, grounded in the teleosemantic distinctions of substance concept and classification concept. We also briefly introduce the ongoing project MultiMedia UKC, whose aim is to build an object recognition resource following our proposed process

Andrea Bontempelli, Fausto Giunchiglia, Andrea Passerini et al.

In this paper, we tackle interactive debugging of "gray-box" concept-based models (CBMs). These models learn task-relevant concepts appearing in the inputs and then compute a prediction by aggregating the concept activations. Our work stems from the observation that in CBMs both the concepts and the aggregation function can be affected by different kinds of bugs, and that fixing these bugs requires different kinds of corrective supervision. To this end, we introduce a simple schema for human supervisors to identify and prioritize bugs in both components, and discuss solution strategies and open problems. We also introduce a novel loss function for debugging the aggregation step that generalizes existing strategies for aligning black-box models to CBMs by making them robust to how the concepts change during training.

Fausto Giunchiglia, Marcelo Rodas Britez, Andrea Bontempelli et al.

The representation of the personal context is complex and essential to improve the help machines can give to humans for making sense of the world, and the help humans can give to machines to improve their efficiency. We aim to design a novel model representation of the personal context and design a learning process for better integration with machine learning. We aim to implement these elements into a modern system architecture focus in real-life environments. Also, we show how our proposal can improve in specifically related work papers. Finally, we are moving forward with a better personal context representation with an improved model, the implementation of the learning process, and the architectural design of these components.

Stefano Teso, Andrea Bontempelli, Fausto Giunchiglia et al.

We tackle sequential learning under label noise in applications where a human supervisor can be queried to relabel suspicious examples. Existing approaches are flawed, in that they only relabel incoming examples that look "suspicious" to the model. As a consequence, those mislabeled examples that elude (or don't undergo) this cleaning step end up tainting the training data and the model with no further chance of being cleaned. We propose Cincer, a novel approach that cleans both new and past data by identifying pairs of mutually incompatible examples. Whenever it detects a suspicious example, Cincer identifies a counter-example in the training set that -- according to the model -- is maximally incompatible with the suspicious example, and asks the annotator to relabel either or both examples, resolving this possible inconsistency. The counter-examples are chosen to be maximally incompatible, so to serve as explanations of the model's suspicion, and highly influential, so to convey as much information as possible if relabeled. Cincer achieves this by leveraging an efficient and robust approximation of influence functions based on the Fisher information matrix (FIM). Our extensive empirical evaluation shows that clarifying the reasons behind the model's suspicions by cleaning the counter-examples helps in acquiring substantially better data and models, especially when paired with our FIM approximation.

Fausto Giunchiglia, Alessio Zamboni, Mayukh Bagchi et al.

We propose a novel approach to the problem of semantic heterogeneity where data are organized into a set of stratified and independent representation layers, namely: conceptual(where a set of unique alinguistic identifiers are connected inside a graph codifying their meaning), language(where sets of synonyms, possibly from multiple languages, annotate concepts), knowledge(in the form of a graph where nodes are entity types and links are properties), and data(in the form of a graph of entities populating the previous knowledge graph). This allows us to state the problem of semantic heterogeneity as a problem of Representation Diversity where the different types of heterogeneity, viz. Conceptual, Language, Knowledge, and Data, are uniformly dealt within each single layer, independently from the others. In this paper we describe the proposed stratified representation of data and the process by which data are first transformed into the target representation, then suitably integrated and then, finally, presented to the user in her preferred format. The proposed framework has been evaluated in various pilot case studies and in a number of industrial data integration problems.

Fausto Giunchiglia, Mayukh Bagchi

We assume that substances in the world are represented by two types of concepts, namely substance concepts and classification concepts, the former instrumental to (visual) perception, the latter to (language based) classification. Based on this distinction, we introduce a general methodology for building lexico-semantic hierarchies of substance concepts, where nodes are annotated with the media, e.g.,videos or photos, from which substance concepts are extracted, and are associated with the corresponding classification concepts. The methodology is based on Ranganathan's original faceted approach, contextualized to the problem of classifying substance concepts. The key novelty is that the hierarchy is built exploiting the visual properties of substance concepts, while the linguistically defined properties of classification concepts are only used to describe substance concepts. The validity of the approach is exemplified by providing some highlights of an ongoing project whose goal is to build a large scale multimedia multilingual concept hierarchy.

Fausto Giunchiglia, Simone Bocca, Mattia Fumagalli et al.

When building a new application we are more and more confronted with the need of reusing and integrating pre-existing knowledge, e.g., ontologies, schemas, data of any kind, from multiple sources. Nevertheless, it is a fact that this prior knowledge is virtually impossible to reuse as-is. This difficulty is the cause of high costs, with the further drawback that the resulting application will again be hardly reusable. It is a negative loop which consistently reinforces itself. iTelos is a general purpose methodology aiming at minimizing as much as possible the effects of this loop. iTelos is based on the intuition that the data level and the schema level of an application should be developed independently, thus allowing for maximum flexibility in the reuse of the prior knowledge, but under the overall guidance of the needs to be satisfied, formalized as competence queries. This intuition is implemented by codifying all the requirements, including those concerning reuse, as part of an a-priori defined purpose, which is then used to drive a middle-out development process where the application schema and data are continuously aligned.

Fausto Giunchiglia, Luca Erculiani, Andrea Passerini

Lexical Semantics is concerned with how words encode mental representations of the world, i.e., concepts . We call this type of concepts, classification concepts . In this paper, we focus on Visual Semantics , namely on how humans build concepts representing what they perceive visually. We call this second type of concepts, substance concepts . As shown in the paper, these two types of concepts are different and, furthermore, the mapping between them is many-to-many. In this paper we provide a theory and an algorithm for how to build substance concepts which are in a one-to-one correspondence with classifications concepts, thus paving the way to the seamless integration between natural language descriptions and visual perception. This work builds upon three main intuitions: (i) substance concepts are modeled as visual objects , namely sequences of similar frames, as perceived in multiple encounters ; (ii) substance concepts are organized into a visual subsumption hierarchy based on the notions of Genus and Differentia ; (iii) the human feedback is exploited not to name objects, but, rather, to align the hierarchy of substance concepts with that of classification concepts. The learning algorithm is implemented for the base case of a hierarchy of depth two. The experiments, though preliminary, show that the algorithm manages to acquire the notions of Genus and Differentia with reasonable accuracy, this despite seeing a small number of examples and receiving supervision on a fraction of them.

Fausto Giunchiglia, Jahna Otterbacher, Styliani Kleanthous et al.

As the role of algorithmic systems and processes increases in society, so does the risk of bias, which can result in discrimination against individuals and social groups. Research on algorithmic bias has exploded in recent years, highlighting both the problems of bias, and the potential solutions, in terms of algorithmic transparency (AT). Transparency is important for facilitating fairness management as well as explainability in algorithms; however, the concept of diversity, and its relationship to bias and transparency, has been largely left out of the discussion. We reflect on the relationship between diversity and bias, arguing that diversity drives the need for transparency. Using a perspective-taking lens, which takes diversity as a given, we propose a conceptual framework to characterize the problem and solution spaces of AT, to aid its application in algorithmic systems. Example cases from three research domains are described using our framework.