Archaeological context is a foundational concept, yet its definition remains ambiguous and inconsistent. Traditionally, lithostratigraphic frameworks rooted in geology have long dominated, but recent computational and spatial approaches increasingly redefine stratigraphic units. This shift raises a critical challenge: the lack of clear and consistent terminology for defining stratigraphic units. To explore this issue, we applied natural language processing (NLP) to a broad corpus of Paleolithic literature, focusing on the use of key terms such as “layer,” “level,” “occupation,” and “horizon.” Using SciBERT (Scientific BERT; BERT = Bidirectional Encoder Representations from Transformers) to generate contextual embeddings, we compared their semantic distributions with a control set of distinct terms (lithic, bone, flake, charcoal). The control set produced stable, separable clusters, whereas contextual terms showed extensive semantic overlap and instability, reflecting their interchangeable and inconsistent use. These results underscore that the problem is not only linguistic but theoretical: without clear terminological baselines, new methodological advances risk conflating analytical tools with interpretive outcomes. We argue that terminological precision is essential for comparability across sites and for cumulative research. NLP offers a replicable framework for auditing and clarifying terminology, contributing to greater rigor and interoperability in archaeological interpretation.

Sentimental analysis (SA) of movie reviews has become an essential means of supporting audiences, filmmakers, and investors to facilitate data-driven marketing, support audience engagement, and maximize audience development returns. Nevertheless, current SA models are yet to overcome potentially daunting problems such as noisy data, ambiguity of language, limited handling of negations, and a loss of rich semantic information during mirror preprocessing that ultimately decreases classification performance. To overcome this hurdle, a unique two-stage strategy entitled Hierarchical Attention-Guided Sentiment Network (HASN) is proposed that enables reliable and interpretable sentiment predictions. The first stage includes a clearly defined preprocessing pipeline of the IMDb dataset that features text cleaning, lemmatizing, removing stop words, negation, TF-IDF filtering of low-frequency terms, and lastly, Word2Vec embeddings, which create semantically consistent and contextually rich representations of potential inputs. The second stage integrates a tri-stream attention recurrent extractor (TARE), which reconstructs and merges previously learned outputs with the learned syntactic and semantic contexts with the benefits of stacked Bi-LSTMs, combines learned syntactic and semantic contexts with multi-head attention to identify the tokens that drive sentiment, and ends with a residual-normalized GRU (ResNorm-GRU), which sequentially refines the sequential process of intake. The attentive conditional random field (Attn-CRF) serves as a classifier by boosting coherence in label predictions through attention weighting with modeling of the sequential dependencies. The experiments with the IMDb dataset demonstrated that the best results detected were the following: 98.09, 98.47, 97.70, and 98.08%. In summary, the HASN framework is a precise, scalable, and interpretable model for analyzing sentiments in movie reviews.

Automatic named entity recognition (NER) is essential for many natural language processing applications, particularly in low-resource languages and those with corpora restricted to a single domain, where the lack of diverse data may hinder cross-domain generalisation. In this context, we attempt to contribute to the Galician NER field by extending manually annotated resources and exploring several approaches to improve the performance of automatic classifiers. To this end, we enrich two new datasets with NER annotations in Galician, which we release in this paper. Through experimental evaluation, we investigate the benefits of fine-tuning transformers on NER data by comparing various monolingual and multilingual BERT models and developing a new extra-small DeBERTa model for Galician. Furthermore, we assess the quality of the NER model by expanding the initial training corpus with additional data from related varieties, including Portuguese and Spanish. For evaluation, we use five Galician corpora, including our two newly annotated datasets and three adapted from existing resources, ensuring assessment across multiple domains. Our findings reveal (1) a positive impact of enhanced cross-lingual training data on improved model performance across five Galician test datasets; (2) similar performance of monolingual and multilingual models depending on the training data; and (3) the competitive effectiveness of extra-small BERT models for Galician NER tasks compared to larger models.

For decades, public opinion scholars have argued for the need to go beyond measuring isolated political preferences to more richly examine how individuals reason about and justify the interconnections between their preferences. While early efforts used interviews and hand-coding to elicit the network structure of subjects’ political reasoning, this manual approach can be time-consuming and difficult. Thankfully, modern computational methods and computing power hold the potential to revolutionize this space and allow for network inference at scale. In this paper, we therefore present a computational, text-based approach for inferring the network structure of individuals’ political reasoning. This method identifies the key concepts a person raises and examines the implicit connections between those concepts—what ideas do they connect to which other ideas? This structural approach is theoretically justified in both the cognitive and linguistic literatures, which repeatedly suggest that humans store, retrieve, and interpret information through network structures. We show that this approach reveals meaningful individual variation that is correlated with known behavioral traits. The ability to measure and interrogate individuals’ expressions of political reasoning holds the potential to shed new light on the dynamics of public opinion and political behavior. Questions of persuasion, ideological fracturing, and conversation quality all rely upon understanding individual styles of political expression. These dynamics are driven not just by what someone says but by how they say it.

Archives are essential vehicles for both historical research and memory practices like genealogy. The sheer volume of documents produced via digital infrastructures, however, has posed a challenge for traditional archival processes that rely on human judgments. Some archivists have responded to this challenge by engaging with research in natural language processing (NLP), a subfield of AI that offers practical tools for working with large corpora of text. In this paper, we review applications of NLP methods in archival processes in order to surface fundamental tensions between archival values and automation. NLP can be used to classify materials as sensitive, recognize names in documents and facilitate new kinds of search, among other applications. Our discussion of these issues follows the journey of a document through an archive, from acquisition to interpretation. We find that sensitivity to context, transparency, resource limitations and maintenance appear as recurring challenges that are complicated by automation. The benefits that NLP technologies lend to managing large archives, however, suggest that these tools should be considered, not discarded. Turning to a more forward-looking discussion of generative AI and archival infrastructure, we argue that the narrative and dialogic potential of contemporary large language models has an especially contentious relationship with conventional archival desiderata.

The practice of public administration is confronted by an overwhelming volume of unstructured textual data, creating a critical bottleneck that challenges the capacity for timely, evidence-based policymaking. While large language models (LLMs) offer a powerful solution, the academic and policy discourse on AI governance has primarily focused on the risks associated with quantitative, predictive models, leaving the governance of LLMs, which involve qualitative, interpretive analysis, critically underexplored. This article provides a systematic and empirical answer to this challenge. We evaluate four distinct archetypes for AI-assisted policy analysis: (1) traditional expert-driven analysis, (2) automated topic modelling, (3) human-in-the-loop LLM co-production (ChatGPT with prompting), and (4) institutional-scale API automation (ChatGPT via API), by applying them to a real-world corpus of 63 United Nations policy documents. Our results reveal the inherent trade-offs of each archetype in terms of analytical speed, cost, contextual depth, and the requisite level of human oversight, finding that the human-in-the-loop model offers the most effective balance. We conclude by presenting a durable framework to guide public institutions in making a strategic, evidence-based choice of analytical workflow, ensuring the adoption of AI strengthens, rather than erodes, the integrity of democratic governance.

Political science increasingly uses text classification to gauge subtle concepts such as toxicity or anger. Traditional methods treat words in isolation, overlooking the contextual dynamics where meaning resides. Transformer-based language models address these limitations but remain underutilized, especially among applied scholars, partly due to misconceptions about their mechanisms and computational requirements. This article bridges this gap by offering an accessible explication of the pretrain-finetune paradigm, focusing on underlying mechanisms and their potential to improve political text analysis by harnessing models trained on extensive datasets while requiring modest labeled data and computational resources. I demonstrate the approach by identifying toxic language in conversation threads following U.S. Senators’ tweets and provide an online tutorial to support nontechnical scholars in adopting the methodology.

Legal epidemiology — the study of how laws influence health outcomes — is an emerging field with potential to inform policy and improve public health. Shortages of specialists who can commit the extensive time required for these analyses has hindered progress in this area. Features of laws make them ideal candidates for artificial intelligence and natural language processing (AI/NLP) to address these constraints: standardized format, defined terms, and common terminology. However, there are concerns regarding valid AI/NLP application, especially given current underreporting of legal research in published policy evaluation studies. Drawing on lessons learned from case deployments and recent literature, we review opportunities and challenges for methods using AI/NLP in scientific legal research: assessing the scope of legal documents and data, identifying and collecting primary legal data, developing and applying coding schemes, and implementing quality controls to ensure the validity of produced legal datasets. We highlight methodologic research areas and innovation in the application of AI/NLP for scientific studies regarding effects of law on health and indicate methodologic reporting elements that will be essential for the field’s innovation. This review provides a method-focused research agenda for AI/NLP in scientific legal epidemiology studies that might accelerate the field’s growth and effects on evidence-based policy.

Skill extraction is at the core of algorithmic hiring. It is based on identifying terms commonly found in both targets (i.e., resumes and job offers), aiming at identifying a “match” or correspondence between both. This article focuses on skill extraction from resumes, as opposed to job offers, and considers this task both from the human resource management (HRM) and artificial intelligence (AI) points of view. We discuss challenges identified by both fields and explain how collaboration is instrumental for a successful digital transformation of HRM. We argue that annotation efforts are an ideal example of where collaboration between both fields is needed and present an annotation effort on 46 resumes with 41 trained annotators, resulting in a total of 116 annotations. We analyze the skills extracted by multiple different systems and compare those to the skills selected by the annotators, and find that the skills extracted differ a lot in terms of length and semantic content. The skills extracted using conversational large language models (LLMs) tend to be very long and detailed; other systems are concise, whereas humans are in the middle. In terms of semantic similarity, conversational LLMs are closer to human outputs than other systems. Our analysis proposes a different perspective to understand the well-studied, but still unsolved, skill extraction task. Finally, we provide policy recommendations for skill extraction aligned with both HRM and computational perspectives.

Suicide is a significant global public health concern, and conventional clinical risk assessment is constrained by workforce availability, scalability, and clinician variability. The classic suicidality risk evaluation is largely dependent on clinical judgment, which, although helpful, can demonstrate a ceiling effect in its predictive validity. AI-driven chatbots, conversational systems that engage users in real-time natural language dialog, have emerged as candidate tools for augmenting suicide risk detection and prevention. This narrative review aimed to: evaluate the performance of AI-driven chatbots in detecting and assessing suicidal ideation relative to clinical benchmarks; examine the effectiveness of chatbot-based interventions for suicide prevention; and identify ethical, cultural, and implementation challenges limiting clinical translation. Six electronic databases were searched, with the initial search conducted in 2024 and updated in 2026 through targeted monitoring, with no upper cutoff date applied. A thematic narrative synthesis approach was applied across five domains. Eleven primary studies met eligibility criteria and were included in the synthesis. Chatbot-based risk assessment showed adequate response alignment with expert judgment at the extremes of suicide risk, but consistently failed to distinguish intermediate risk levels across multiple model families. Across 29 tested commercial chatbot agents, none met the criteria for an adequate suicidal crisis response. A clinically designed, framework-anchored chatbot achieved high efficacy across six outcome domains. Three percent of social chatbot users reported halted suicidal ideation, and a purpose-built clinical chatbot in emergency department settings significantly improved evidence-based care delivery. Systematic risk severity underestimation and the absence of cross-cultural evaluation were identified across all studies. AI-driven chatbots show potential as adjunctive tools across the suicide care continuum. Clinically designed, evidence-anchored chatbots demonstrate feasibility and meaningful benefit; commercially deployed chatbots without clinical validation demonstrate near-universal crisis response inadequacy. Mandatory clinical validation prior to public release, clinician oversight, and crisis system integration are prerequisites for responsible deployment.

Gauging the extent of public acceptability of reforms is an important concern for policymakers. Timely insights into public perceptions can illuminate how reforms are received and how attitudes evolve over time. In this study, we build on the OECD’s Public Acceptability Tool, a framework encompassing four key dimensions of reform acceptability—Economic, Fairness, Behavioural, and Process—to evaluate the public acceptability of policy reforms. We take the 2023 French pension reform as a relevant case study, using online media articles and parliamentary speeches as indicators of discourse surrounding the reform. Using word embeddings, we classify these texts according to the four dimensions and apply matrix factorisation topic algorithms to uncover the latent themes within each. Our analysis shows that the Process dimension dominated media coverage during the discussion and legislative phases of the reform, consistent with previous literature on pension reforms. In contrast, no particular dimension was predominant in parliamentary speeches, suggesting a mismatch between policy and public debates. Finally, we identify the main topics driving public discussion within each dimension, highlighting notable differences between media narratives and parliamentary discourse that offer further insight into the dynamics of public acceptability.