1,720,972 research outputs found
SPARQL queries over source code
No full textWe introduce a framework to extract and parse Java source code, serialize it into RDF triples by applying an appropriate ontology and then analyze the resulting structured code information by using standard SPARQL queries. We present our experiments on a sample of 134 Java repositories collected from Github, obtaining 17 Million triples about methods, input and output types, comments, and other source code information. Experiments also address the scalability of the framework. We finally provide examples of the level of expressivity that can be achieved with SPARQL by using our proposed ontology and semantic technologies
Explainable authorship identification in cultural heritage applications
Full text linkWhile a substantial amount of work has recently been devoted to improving the accuracy of computational Authorship Identification (AId) systems for textual data, little to no attention has been paid to endowing AId systems with the ability to explain the reasons behind their predictions. This substantially hinders the practical application of AId methods, since the predictions returned by such systems are hardly useful unless they are supported by suitable explanations. In this article, we explore the applicability of existing general-purpose eXplainable Artificial Intelligence (XAI) techniques to AId, with a focus on explanations addressed to scholars working in cultural heritage. In particular, we assess the relative merits of three different types of XAI techniques (feature ranking, probing, factual and counterfactual selection) on three different AId tasks (authorship attribution, authorship verification and same-authorship verification) by running experiments on real AId textual data. Our analysis shows that, while these techniques make important first steps towards XAI, more work remains to be done to provide tools that can be profitably integrated into the workflows of scholars
FairBelief - Assessing Harmful Beliefs in Language Models
No full textLanguage Models (LMs) have been shown to inherit undesired biases that might hurt minorities and underrepresented groups if such systems were integrated into real-world applications without careful fairness auditing.This paper proposes FairBelief, an analytical approach to capture and assess beliefs, i.e., propositions that an LM may embed with different degrees of confidence and that covertly influence its predictions. With FairBelief, we leverage prompting to study the behavior of several state-of-the-art LMs across different previously neglected axes, such as model scale and likelihood, assessing predictions on a fairness dataset specifically designed to quantify LMs’ outputs’ hurtfulness.Finally, we conclude with an in-depth qualitative assessment of the beliefs emitted by the models.We apply FairBelief to English LMs, revealing that, although these architectures enable high performances on diverse natural language processing tasks, they show hurtful beliefs about specific genders. Interestingly, training procedure and dataset, model scale, and architecture induce beliefs of different degrees of hurtfulness
Towards Synergistic Human-AI Collaboration in Hybrid Decision-Making Systems
No full textA growing body of interdisciplinary literature indicates that human decision-making processes can be enhanced by Artificial Intelligence (AI). Nevertheless, the use of AI in critical domains has also raised significant concerns regarding its final users, those affected by the undertaken decisions, and the broader society. Consequently, recent studies are shifting their focus towards the development of human-centered frameworks that facilitate a synergistic human-machine collaboration while upholding ethical and legal standards. In this work, we present a taxonomy for hybrid decision-making systems to classify systems according to the type of interaction that occurs between human and artificial intelligence. Furthermore, we identify gaps in the current body of literature and suggest potential directions for future research
Generative Model for Decision Trees
Full text linkDecision trees are among the most popular supervised models due to their interpretability and knowledge representation resembling human reasoning. Commonly-used decision tree induction algorithms are based on greedy top-down strategies.
Although these approaches are known to be an efficient heuristic, the resulting trees are only locally optimal and tend to have overly complex structures. On the other hand, optimal decision tree algorithms attempt to create an entire decision tree at once to achieve global optimality. We place our proposal between these approaches by designing a generative model for decision trees. Our method first learns a latent decision tree space through a variational architecture using pre-trained decision tree models. Then, it adopts a genetic procedure to explore such latent space to find a compact decision tree with good predictive performance. We compare our proposal against classical tree induction methods, optimal approaches, and ensemble models. The results show that our proposal can generate accurate and shallow, i.e., interpretable, decision trees
GLocalX - From Local to Global Explanations of Black Box AI Models
Full text linkArtificial Intelligence (AI) has come to prominence as one of the major components of our society, with applications in most aspects of our lives. In this field, complex and highly nonlinear machine learning models such as ensemble models, deep neural networks, and Support Vector Machines have consistently shown remarkable accuracy in solving complex tasks. Although accurate, AI models often are “black boxes” which we are not able to understand. Relying on these models has a multifaceted impact and raises significant concerns about their transparency. Applications in sensitive and critical domains are a strong motivational factor in trying to understand the behavior of black boxes. We propose to address this issue by providing an interpretable layer on top of black box models by aggregating “local” explanations. We present GLocalX, a “local-first” model agnostic explanation method. Starting from local explanations expressed in form of local decision rules, GLocalX iteratively generalizes them into global explanations by hierarchically aggregating them. Our goal is to learn accurate yet simple interpretable models to emulate the given black box, and, if possible, replace it entirely. We validate GLocalX in a set of experiments in standard and constrained settings with limited or no access to either data or local explanations. Experiments show that GLocalX is able to accurately emulate several models with simple and small models, reaching state-of-the-art performance against natively global solutions. Our findings show how it is often possible to achieve a high level of both accuracy and comprehensibility of classification models, even in complex domains with high-dimensional data, without necessarily trading one property for the other. This is a key requirement for a trustworthy AI, necessary for adoption in high-stakes decision making applications
HANSEN: Human and AI Spoken Text Benchmark for Authorship Analysis
Full text linkAuthorship Analysis, also known as stylometry, has been an essential aspect of Natural Language Processing (NLP) for a long time. Likewise, the recent advancement of Large Language Models (LLMs) has made authorship analysis increasingly crucial for distinguishing between human-written and AI-generated texts. However, these authorship analysis tasks have primarily been focused on written texts, not considering spoken texts. Thus, we introduce the largest benchmark for spoken texts - HANSEN (Human ANd ai Spoken tExt beNchmark).
HANSEN encompasses meticulous curation of existing speech datasets accompanied by transcripts, alongside the creation of novel AI-generated spoken text datasets. Together, it comprises 17 human datasets, and AI-generated spoken texts created using 3 prominent LLMs: ChatGPT, PaLM2, and Vicuna13B. To evaluate and demonstrate the utility of HANSEN, we perform Authorship Attribution (AA) & Author Verification (AV) on human-spoken datasets and conducted Human vs. AI spoken text detection using state-of-the-art (SOTA) models. While SOTA methods, such as, character ngram or Transformer-based model, exhibit similar AA & AV performance in human-spoken datasets compared to written ones, there is much room for improvement in AI-generated spoken text detection. The HANSEN benchmark is available at: https://huggingface.co/datasets/HANSEN-REPO/HANSEN
- …
