Abstract
Story components, namely, events, time, participants, and their relations are present in narrative texts from different domains such as journalism, medicine, finance, and law. The automatic extraction of narrative elements encompasses several NLP tasks such as Named Entity Recognition, Semantic Role Labeling, Event Extraction, and Temporal Inference. The text2story Python, an easy-to-use modular library, supports the narrative extraction and visualization pipeline. The package contains an array of narrative extraction tools that can be used separately or in sequence. With this toolkit, end users can process free text in English or Portuguese and obtain formal representations, like standard annotation files or a formal logical representation. The toolkit also enables narrative visualization as Message Sequence Charts (MSC), Knowledge Graphs, and Bubble Diagrams, making it useful to visualize and transform human-annotated narratives. The package combines the use of off-the-shelf and custom tools and is easily patched (replacing existing components) and extended (e.g. with new visualizations). It includes an experimental module for narrative element effectiveness assessment and being is therefore also a valuable asset for researchers developing solutions for narrative extraction. To evaluate the baseline components, we present some results of the main annotators embedded in our package for datasets in English and Portuguese. We also compare the results with the extraction of narrative elements by GPT-3, a robust LLM model.

Abstract

Abstract
E-commerce has become an essential aspect of modern life, providing consumers globally with convenience and accessibility. However, the high volume of short and noisy product descriptions in text streams of massive e-commerce platforms translates into an increased number of clusters, presenting challenges for standard model-based stream clustering algorithms. Standard LDA-based methods often lead to clusters dominated by single elements, effectively failing to manage datasets with varied cluster sizes. Our proposed Community-Based Topic Modeling with Contextual Outlier Handling (CB-TMCOH) algorithm introduces an approach to outlier detection in text data using transformer models for similarity calculations and graph-based clustering. This method efficiently separates outliers and improves clustering in large text datasets, demonstrating its utility not only in e-commerce applications but also proving effective for news and tweets datasets.

Abstract
The proposed framework (Human-Centered Trustworthy Framework) provides a novel human-computer interaction approach to incorporate positive and meaningful trustful user experiences in the system design process. It helps to illustrate potential users' trust concerns in artificial intelligence and guides nonexperts to avoid designing vulnerable interactions that lead to breaches of trust.

Abstract
Aims. The goal of this work is to characterize the polarization effects of the beam path of the Very Large Telescope Interferometer (VLTI) and the GRAVITY beam combiner instrument. This is useful for two reasons: to calibrate polarimetric observations with GRAVITY for instrumental effects and to understand the systematic error introduced to the astrometry due to birefringence when observing targets with a significant intrinsic polarization. Methods. By combining a model of the VLTI light path and its mirrors and dedicated experimental data, we constructed a full polarization model of the VLTI Unit Telescopes (UTs) and the GRAVITY instrument. We first characterized all telescopes together to construct a universal UT calibration model for polarized targets with the VLTI. We then expanded the model to include the differential birefringence between the UTs. With this, we were able to constrain the systematic errors and the contrast loss for highly polarized targets. Results. Along with this paper, we have published a standalone Python package that can be used to calibrate the instrumental effects on polarimetric observations. This enables the community to use GRAVITY with the UTs to observe targets in a polarimetric observing mode. We demonstrate the calibration model with the Galactic Center star IRS 16C. For this source, we were able to constrain the polarization degree to within 0.4% and the polarization angle to within 5 degrees while being consistent with the literature values. Furthermore, we show that there is no significant contrast loss, even if the science and fringe-tracker targets have significantly different polarization, and we determine that the phase error in such an observation is smaller than 1 degrees, corresponding to an astrometric error of 10 mu as. Conclusions. With this work, we enable the use by the community of the polarimetric mode with GRAVITY/UTs and outline the steps necessary to observe and calibrate polarized targets with GRAVITY. We demonstrate that it is possible to measure the intrinsic polarization of astrophysical sources with high precision and that polarization effects do not limit astrometric observations of polarized targets.

Abstract
Estimating turbulence parameters is essential during commissioning and optimising adaptive optics or fringe tracking systems. It also gained new relevance with free-space optical communication applications. The estimation of such parameters is done under the assumption of stationarity. Yet, the stationarity time scale of the atmospheric turbulence is unknown. The breakdown of this assumption leads to incorrect estimates and added error terms. In this paper, we illustrate stationarity detection with unit root testing and the pitfalls of its application to turbulence parameter time series.