Contemporary Arabic narrative generation faces a critical cultural authenticity crisis, where neural models systematically violate classical adab principles despite sophisticated linguistic capabilities. We introduce CST-Onto, a hybrid neuro-symbolic framework addressing this challenge through the first domain-specific OWL/SWRL ontology for Arabic cultural narratives. Our four-tier ontology encodes 87 core cultural archetypes from expert annotations of Kalīla wa Dimna and One Thousand and One Nights [13], expanded to 847 specialized concepts through systematic derivation protocols. The framework formalizes 234 SWRL rules modeling virtue dependencies (e.g., ḥakīm → ḥilm, ʿadl, faṣāḥa) that directly enhance symbolic–neural integration via constraint-aware neural decoding.

Arabic text adaptation confronts severe computational barriers from morphological complexity and dialectal fragmentation. Current graph-based methods employ static representations that fail to capture evolving morphological dependencies, causing substantial semantic degradation (28.0% cross-dialectal error rate) across dialectal boundaries. We introduce MM-GTA (Morphological Modeling via Graph-based Text Adaptation), the first dynamic-graph framework for Arabic style transfer that jointly optimizes morphological dynamism, dialect-conditioned attention, and cultural-adaptive resource fusion (dialect-aware embedding compression). MM-GTA introduces dynamic graph reconstruction and dialect-conditioned attention, reducing cross-dialect errors by 55.8% versus AraT5 (from 28.0% to 12.3%). Evaluation on ArabicStyleBank (10,000 parallel texts across five dialects) demonstrates substantial gains: BLEU 41.2±0.8 (+6.5 over AraBERT’s 34.7±0.9), ROUGE-L 71.8±1.0 (+6.1 points), BERTScore 78.2±1.0 (+4.3 points). The architecture achieves 51% memory reduction with 89ms average inference latency per sentence (vs. AraT5’s 142ms), enabling efficient deployment within 8.0GB memory constraints. Human evaluation by 25 native speakers (5 per dialect, avg. 10 years NLP experience) confirms superior preservation across style authenticity (4.2/5), semantic fidelity (4.3/5), and grammatical accuracy (4.1/5). Our model achieves +17.2% improvement in cultural alignment compared to AraGPT2 across three major dialect families.

Suspense is one key feature in modern storytelling. One of the mechanisms to deliver a suspenseful experience to an audience is by means of controlling the information provided. The media, however, has a very strong impact on what kind of information can be delivered and how. Moreover, modern storytelling is usually conveyed interactively, in such a way that the audience is also part of the story. In this paper, we experiment and analyse the different impact of information management in interactive and non-interactive storytelling. We report on an experiment measuring the reported perceived amusement in interactive and non-interactive versions of a potentially suspenseful story, and we provide evidence that a passive, non-interactive audience usually prefers less information than an active interactive audience. The study provides informed insight on how these results could be used in real scenarios to deliver appropriate levels of information to enhance the perception of suspense.

Interactive storytelling systems have become very popular as they engage users in the creation of narrative. A fundamental challenge for such systems is that the users feel unconstrained in their exploration of the environment and yet retain for the author some control of what the user does. Traditional solutions address this challenge by hardening the authoring task, since much effort has to be devoted to create the story and the necessary mechanisms to provide the user with enough freedom that, at the same time, keeps the story consistent. We propose a solution based on dynamic narration that combines a set of authored materials associated with particular points in the environment and the exploratory movements of the user, and includes statements of varying emphasis to guide the user towards the desired elements. This allows the system to generate a narration in real time while giving the user the possibility to decide what to do. The solution has been field tested and results are reported on three different aspects: the freedom of the user in terms of combinations in which the authored material is traversed; the divergence of user explorations from author's intentions; and user response to guidance statements of differing emphasis.

This paper addresses the task of generating descriptions of people for an observer that is moving within a scene.As the observer moves, the descriptions of the people around him also change.A referring expression generation algorithm adapted to this task needs to continuously monitor the changes in the field of view of the observer, his relative position to the people being described, and the relative position of these people to any landmarks around them, and to take these changes into account in the referring expressions generated.This task presents two advantages: many of the mechanisms already available for static contexts may be applied with small adaptations, and it introduces the concept of changing conditions into the task of referring expression generation.In this paper we describe the design of an algorithm that takes these aspects into account in order to create descriptions of people within a 3D virtual environment.The evaluation of this algorithm has shown that, by changing the descriptions in real time according to the observers point of view, they are able to identify the described person quickly and effectively.

The article explores the potential of redefining as services a number of functionalities involved in poetry generation systems to better serve the challenge of working toward a yet unknown successful computational model of the creative task being addressed. This is performed by considering how some of the processes currently modeled in existing systems for poetry generation might be deconstructed into a set of services susceptible of being recombined in different ways to be integrated in other developments. Building on prior attempts to propose an evolutionary architecture that allows integration of a number of artificial intelligence technologies in combination, this article explores the advantages of service‐oriented architecture for reimplementing as publicly available services some selected functionality of the Wishful Automatic Spanish Poet (WASP) poetry generation system.

Suspense is a key narrative issue in terms of emotional gratification, influencing the way in which the audience experiences a story. Virtually all narrative media uses suspense as a strategy for reader engagement regardless of the technology or genre. Being such an important narrative component, computational creativity has tackled suspense in a number of automatic storytelling. These systems are mainly based on narrative theories, and in general lack a cognitive approach involving the study of empathy or emotional effect of the environment impact. With this idea in mind, this paper reports on a computational model of the influence of decorative elements on suspense. It has been developed as part of a more general proposal for plot generation based on cognitive aspects. In order to test and parameterise the model, an evaluation based on textual stories and an evaluation based on a 3D virtual environment was run. In both cases, results suggest a direct influence of emotional perception of decorative objects in the suspense of a scene.

Abstract Poetry generation is a specific kind of natural language generation where several sources of knowledge are typically exploited to handle features on different levels, such as syntax, semantics, form or aesthetics. But although this task has been addressed by several researchers, and targeted different languages, all known systems have focused on a limited purpose and a single language. This article describes the effort of adapting the same architecture to generate poetry in three different languages – Portuguese, Spanish and English. An existing architecture is first described and complemented with the adaptations required for each language, including the linguistic resources used for handling morphology, syntax, semantics and metric scansion. An automatic evaluation was designed in such a way that it would be applicable to the target languages. It covered three relevant aspects of the generated poems, namely: the presence of poetic features, the variation of the linguistic structure and the semantic connection to a given topic. The automatic measures applied for the second and third aspect can be seen as novel in the evaluation of poetry. Overall, poems were successfully generated in the three languages addressed. Despite minor differences in different languages or seed words, poems revealed to have a regular metre, frequent rhymes, to exhibit an interesting degree of variation, and to be semantically-associated with the initially given seeds.

Most poetry-generation systems apply opportunistic approaches where algorithmic procedures are applied to explore the conceptual space defined by a given knowledge resource in search of solutions that might be aesthetically valuable. Aesthetical value is assumed to arise from compliance to a given poetic form – such as rhyme or metrical regularity – or from evidence of semantic relations between the words in the resulting poems that can be interpreted as rhetorical tropes – such as similes, analogies, or metaphors. This approach tends to fix a priori the aesthetic parameters of the results, and imposes no constraints on the message to be conveyed. The present paper describes an attempt to initiate a shift in this balance, introducing means for constraining the output to certain topics and allowing a looser mechanism for constraining form. This goal arose as a result of the need to produce poems for a themed collection commissioned to be included in a book. The solution adopted explores an approach to creativity where the goals are not solely aesthetic and where the results may be surprising in their poetic form. An existing computer poet, originally developed to produce poems in a given form but with no specific constraints on their content, is put to the task of producing a set of poems with explicit restrictions on content, and allowing for an exploration of poetic form. Alternative generation methods are devised to overcome the difficulties, and the various insights arising from these new methods and the impact they have on the set of resulting poems are discussed in terms of their potential contribution to better poetry-generation systems.

Abstract Variability is inherent in human language as different people make different choices when facing the same communicative act. In Natural Language Processing, variability is a challenge. It hinders some tasks such as evaluation of generated expressions, while it constitutes an interesting resource to achieve naturalness and to avoid repetitiveness. In this work, we present a methodological approach to study the influence of lexical variability. We apply this approach to TUNA, a corpus of referring expression lexicalizations, in order to study the use of different lexical choices. First, we reannotate the TUNA corpus with new information about lexicalization, and then we analyze this reannotation to study how people lexicalize referring expressions. The results show that people tend to be consistent when generating referring expressions. But at the same time, different people also share certain preferences.

This paper presents an approach to the automated markup of texts with emotional labels. The approach considers two possible representations of emotions in parallel: emotional categories (emotional tags used to refer to emotions) and emotional dimensions (measures that try to model the essential aspects of emotions numerically). For each representation, a corpus of example texts previously annotated by human evaluators is mined for an initial assignment of emotional features to words. This results in a list of emotional words (LEW) which becomes a useful resource for later automated markup. The algorithm proposed for the automated markup of text closely mirrors the steps taken during feature extraction, employing a combination of the LEW resource and the ANEW word list for the actual assignment of emotional features, and WordNet for knowledge‐based expansion of words not occurring in either and an ontology of emotional categories. The algorithm for automated markup is tested and the results are discussed with respect to three main issues: the relative adequacy of each of the representations used, correctness and coverage of the proposed algorithm, and additional techniques and solutions that may be employed to improve the results. The average percentage of success obtained by our approach when it marks up with emotional dimensions is around 80% and when it marks up with emotional categories is around 50%. The main contribution of the approach presented in this paper is that it allows dimensions and categories at different levels of abstraction to operate simultaneously during markup.

Assisted story generation systems do not include automatic generation of content due to the fact that creating generation algorithms is a challenging task, usually carried out in research projects on Artificial Intelligence. This paper proposes a top-down design methodology in which the knowledge representation is based on causality and chronology. The proposed methodology partially eases the process by dividing the narrative generation system in two parts: a generic engine and a domain definition based only on a specific set of predicates. The theoretical model and an implemented case study with resulting stories are presented.