Large Language Models (LLMs) represent a promising frontier for recommender systems, yet their development has been impeded by the absence of predictable scaling laws, which are crucial for guiding research and optimizing resource allocation. We hypothesize that this may be attributed to the inherent noise, bias, and incompleteness of raw user interaction data in prior continual pre-training (CPT) efforts. This paper introduces a novel, layered framework for generating high-quality synthetic data that circumvents such issues by creating a curated, pedagogical curriculum for the LLM. We provide powerful, direct evidence for the utility of our curriculum by showing that standard sequential models trained on our principled synthetic data significantly outperform ($+130\%$ on recall@100 for SasRec) models trained on real data in downstream ranking tasks, demonstrating its superiority for learning generalizable user preference patterns. Building on this, we empirically demonstrate, for the first time, robust power-law scaling for an LLM that is continually pre-trained on our high-quality, recommendation-specific data. Our experiments reveal consistent and predictable perplexity reduction across multiple synthetic data modalities. These findings establish a foundational methodology for reliable scaling LLM capabilities in the recommendation domain, thereby shifting the research focus from mitigating data deficiencies to leveraging high-quality, structured information.
In an era of globalised security threats – from terrorism and cyber-attacks to hybrid warfare and organised crime – European intelligence services increasingly rely on intensive international cooperation. New technologies enable the rapid and extensive sharing of sensitive data, while the transnational nature of threats forces agencies to cross national borders and build common networks, databases and coordination mechanisms. This cooperation has deepened significantly in recent years, making the European intelligence space one of the most integrated in the world. Paradoxically, this dynamic development of cooperation in the field of information gathering and exchange is contrasted by an almost complete absence of parallel progress in international oversight of intelligence services. Control and oversight remain largely a national domain, constrained by strict national rules and limited capacity to scrutinise cross-border activities. This asymmetric development raises fundamental questions about democratic legitimacy, the protection of fundamental rights and accountability in an era of massive intelligence sharing. This article addresses this imbalance, analysing the current state of European intelligence cooperation, mapping key institutions and information-sharing mechanisms, and, above all, examining weaknesses in transnational oversight. The aim is to highlight the risks stemming from the absence of effective international oversight mechanisms and to suggest possible ways to strengthen them in order to ensure a balance between security needs and the protection of democratic principles in Europe.
With the acceleration of urbanization, criminal behavior in public scenes poses an increasingly serious threat to social security. Traditional anomaly detection methods based on feature recognition struggle to capture high-level behavioral semantics from historical information, while generative approaches based on Large Language Models (LLMs) often fail to meet real-time requirements. To address these challenges, we propose MA-CBP, a criminal behavior prediction framework based on multi-agent asynchronous collaboration. This framework transforms real-time video streams into frame-level semantic descriptions, constructs causally consistent historical summaries, and fuses adjacent image frames to perform joint reasoning over long- and short-term contexts. The resulting behavioral decisions include key elements such as event subjects, locations, and causes, enabling early warning of potential criminal activity. In addition, we construct a high-quality criminal behavior dataset that provides multi-scale language supervision, including frame-level, summary-level, and event-level semantic annotations. Experimental results demonstrate that our method achieves superior performance on multiple datasets and offers a promising solution for risk warning in urban public safety scenarios.
In criminal justice, forensic autopsies play a critical role in determining the cause and manner of death. Current provisions are unclear, leading to inconsistencies in the criminal investigation process. This study examines weaknesses in Indonesian law, particularly the ambiguity in regulating forensic autopsies, the obligation to conduct autopsies in suspected criminal deaths, and the procedures for notifying the victim's family. This study aims to provide an interpretation of the legal regulations concerning autopsies and to formulate regulatory reforms related to the conduct of autopsies in criminal investigation processes. The results of this study indicate that in criminal investigations, autopsies are crucial for determining the cause and manner of death, as well as providing scientific evidence linking the perpetrator's actions to the victim's death. However, ambiguity in interpreting the obligation to conduct autopsies under Article 134 of the Criminal Procedure Code (KUHAP) leads to inconsistencies. To address this, clear regulations are needed to define when an autopsy is considered "very necessary," providing consistent guidance to investigators. Mandatory autopsies for all unnatural deaths would ensure thorough examinations. While families must be informed, their consent is not required, and obstructing an autopsy is a criminal offence, supported by Article 222 of the Criminal Code, as amended by Article 283 of Law No. 1 of 2023 concerning the Criminal Code.
T. Slipeniuk, M. Yankovyi, Viktor Nikitenko
et al.
Objective: The object of this scientific publication is the use of electronic evidence in criminal proceedings, focusing on the main problematic issues, legislative gaps, and the need for improvements in law enforcement practices. Method: The method of study involves a detailed analysis of the current criminal procedure legislation, identification of the general and special features of electronic evidence, and examination of practical issues related to the collection, assessment, and admissibility of electronic evidence in criminal proceedings. The authors use legal analysis and case studies to propose amendments to the legislation and suggest practical improvements. Results: In this scientific publication, the authors highlight the main problematic issues of using electronic evidence in criminal proceedings, propose amendments to the current criminal procedure legislation, and also make proposals for improving law enforcement practice. It is argued that electronic evidence is an independent procedural source of evidence which has both general and special features and nowadays plays an important role in establishing the truth in criminal proceedings. It is concluded that the use of electronic evidence in criminal proceedings currently has no regulatory basis, which causes numerous difficulties and gaps which, among other things, lead to abuse of procedural powers. Contribution: It is established that in the process of collecting electronic evidence, there are many abuses of procedural powers, which are expressed in the excessive seizure of their carriers, as well as unlawful interference with privacy. The article also reveals the difficulties in assessing electronic evidence as relevant, admissible and reliable evidence. In particular, the use of different copies of electronic documents, compliance by judges with the principle of direct examination of evidence, and the admissibility of electronic evidence obtained from open sources are controversial. Based on the results of the study, the authors propose to amend the Code of Criminal Procedure of Ukraine by supplementing Chapter 4 ‘Evidence and Proof’ with paragraph 5 ‘Electronic Evidence’, which should define their concept, procedural procedure for their seizure, storage, return, and peculiarities of evaluation. The authors also emphasises the expediency of upgrading the qualifications of authorised persons by conducting thematic training on the proper use of electronic evidence and preventing abuse of procedural powers in this process.
This research discusses that the application of discretion often causes problems. However, on the one hand, discretion is also helpful for the Police in carrying out the investigation and investigation process so that they can quickly solve a problem that arises in the field. Law Number 2 of 2002 concerning the National Police of the Republic of Indonesia was born to try to put discretion into the practice of criminal law. The Criminal Procedure Code describes how discretion is applied, but the measure of the application of discretion, whether this discretion is better to exist or not, the value of discretion applied in Indonesia, and how we can exercise discretion. It has yet to be answered, and it is often a problem in the field because it seems arbitrary; for this reason, it is necessary to revise Law Number 2 of 2002 concerning the Police, which explains what discretion is and how it is applied. The research method used is normative legal research, which traces the legal literature linked to the discretionary concept approach presented by Hart.
ex-ante judicial supervision over prosecution and investigation is one of the new developments in criminal procedure. This supervision is the use of supervision by court judges or independent expert judges outside of the Prosecutor's Office over the actions and decisions of the Prosecutor's Office, including the investigator, especially actions that impact individual rights and freedoms in the pre-trial stage of criminal proceedings. Ex-ante supervision is recognized in the criminal justice system of Italy, and Germany; however, in Iran's law, ex-ante judicial supervision has been considered only case-by-case in some articles of Iran's Criminal Procedure Code, without being correctly extended to all decisions that are related to the rights and freedoms of individuals in the pre-trial stage. This article aims to explain the ex-ante judicial supervision comparatively taken in Germany and Italy and examine their differences with the existing supervision in Iran using analytical and descriptive methods, and at the same time, it is comparative with library collection tools; done. The findings of the research indicate the need for Iranian legislators and legal doctrine to pay attention to the issue of ex-ante judicial supervision over prosecution and investigation, such as in Germany and Italy, in an explicit and precise manner, and to prevent the centralization of powers of prosecution, investigation and supervision in the hands of the prosecutor's office to improve the quality of criminal justice and Prevention of judicial errors and violation of people's rights with these broad powers that were given to the judiciary.
Pernille Lund Hansen, Peter Mygind Leth, Pernille Aagaard Nielsen
et al.
Aim: Postmortem Computed Tomography (PMCT) is gradually introduced at forensic institutes. Image reconstruction software can increase diagnostic potential in CT by increasing distinction between structures and reduction of artifacts. The aim of this study was to develop and evaluate novel image reconstruction parameters for postmortem conditions, to increase image quality and diagnostic potential of CT scans. Method: Twenty PMCT scans of deceased hereof two in severe decay were subjected to four reconstruction techniques: a standard reconstruction algorithm, the detail reconstruction algorithm and two novel algorithms based on the standard algorithm, but with different Hounsfield settings. Image quality was evaluated by visual grading analysis (VGA) by four forensic radiologist observers. Results: The VGA did not prove that any of the reconstruction techniques were superior to the others. For standard and detail, the two pre-defined reconstruction algorithms, VGA scores were indiscernible and were superior to the equally indiscernible Hounsfield reconstructions on parameters translated into Sharpness and Low Contrast Resolution. The two alternative Hounsfield settings were superior with respect to Noise and Artifacts/Beam Hardening. Conclusion: The study elucidates the possiblity for multiple reconstructions specialized for PMCT conditions, to accommodate the special conditions when working with the deceased. Despite the lack of clear improvements in the tested reconstructions, this study provides an insight into some of the possibilities of improving PMCT quality using reconstruction techniques.
The effectiveness of Large Language Models (LLMs) in legal reasoning is often limited due to the unique legal terminologies and the necessity for highly specialized knowledge. These limitations highlight the need for high-quality data tailored for complex legal reasoning tasks. This paper introduces LegalSemi, a benchmark specifically curated for legal scenario analysis. LegalSemi comprises 54 legal scenarios, each rigorously annotated by legal experts, based on the comprehensive IRAC (Issue, Rule, Application, Conclusion) framework from Malaysian Contract Law. In addition, LegalSemi is accompanied by a structured knowledge base (SKE). A series of experiments were conducted to assess the usefulness of LegalSemi for IRAC analysis. The experimental results demonstrate the effectiveness of incorporating the SKE for issue identification, rule retrieval, application and conclusion generation using four different LLMs.
Julie Marcotte, Marie-Claude Richard, Isabelle F.-Dufour
et al.
Les objectifs du présent article sont de mettre en exergue la façon dont les jeunes décrivent leur histoire traumatique et lui donnent un sens, et comment ils intègrent ces éléments dans leur perception de soi et leur vision d’avenir. Trente et un récits de vie de jeunes ayant connu une histoire de placement en vertu de la protection de la jeunesse ont été analysés en s’appuyant sur les prémisses de l’analyse thématique théorique (ATT) (Boyatzis, 1998) et à partir des théories du trauma complexe et l’identité narrative. Les résultats montrent que les jeunes sont en mesure de comprendre et décrire les effets des traumatismes vécus sur leur construction de soi, de donner un sens aux stratégies utilisées pour faire face à ces évènements et la manière dont ils façonnent leur avenir. Parce que le récit de vie permet de donner un sens à l’expérience traumatique, la discussion aborde la pertinence de l’utiliser et aussi, plus largement, l’approche narrative, pour bien saisir les enjeux complexes que vivent les jeunes placés dans leur construction identitaire et pour soutenir l’intervention auprès d’eux alors qu’ils amorcent leur transition vers la vie adulte.
Vesna Ćorić, Ana Knežević Bojović, Milica V. Matijević
Currently, the field of business and human rights is at a crossroads in terms of normative development, as two major legislative instruments are being negotiated at the regional and international levels. The first instrument is a proposal for a directive aimed at ensuring business responsibility for the respect of human rights and the environment within the European Union, or in other words a proposal for a Directive on Corporate Sustainability Due Diligence. The second one is a proposal of a legally binding instrument on transnational corporations and other business enterprises with respect to human rights, commonly referred to as the Third Revised Draft Treaty on Business and Human Rights, which is being developed by the open-ended intergovernmental working group established by the Human Rights Council in 2014. Given such parallel developments, it would seem prudent for the ongoing efforts to be interlinked so as to contribute to creating consistent legal solutions governing corporate accountability for human rights violations at international and supranational fora. This is particularly relevant in the context of rapid globalization, where transnational corporations can exploit legal and regulatory loopholes at the cost of human rights and the environment. This paper analyses the two legislative drafts with the aim of determining to what extent those two draft hard law instruments reflect the applicable international soft law standards and contribute to the creation of a complementary and mutually reinforcing regulatory framework. The analysis shows the differences in the scope and approaches utilized in the two instruments and identifies gaps and shortcomings in the proposed solutions from the standpoint of effective protection of the victims’ rights. The analysis shows that the two proposed legislative texts are for the most part mutually complementary and it points to the ways in which their norms can be read together so as to enable a coherent and consistent legal framework and ensure legal certainty. The authors also argue that the two legislators should utilize the drafting process to address the identified discrepancies in the existing normative framework in order to achieve the best results.
Private international law. Conflict of laws, Criminal law and procedure
Machine learning based decision-support tools in criminal justice systems are subjects of intense discussions and academic research. There are important open questions about the utility and fairness of such tools. Academic researchers often rely on a few small datasets that are not sufficient to empirically study various real-world aspects of these questions. In this paper, we contribute WCLD, a curated large dataset of 1.5 million criminal cases from circuit courts in the U.S. state of Wisconsin. We used reliable public data from 1970 to 2020 to curate attributes like prior criminal counts and recidivism outcomes. The dataset contains large number of samples from five racial groups, in addition to information like sex and age (at judgment and first offense). Other attributes in this dataset include neighborhood characteristics obtained from census data, detailed types of offense, charge severity, case decisions, sentence lengths, year of filing etc. We also provide pseudo-identifiers for judge, county and zipcode. The dataset will not only enable researchers to more rigorously study algorithmic fairness in the context of criminal justice, but also relate algorithmic challenges with various systemic issues. We also discuss in detail the process of constructing the dataset and provide a datasheet. The WCLD dataset is available at \url{https://clezdata.github.io/wcld/}.
California experienced an increase in violent criminality during the last decade, largely driven by a surge in aggravated assaults. To address this challenge, accurate and timely forecasts of criminal activity may help state authorities plan ahead and distribute public resources efficiently to reduce crime. This paper forecasts monthly aggravated assaults in California using a publicly available dataset on state crimes and a time series SARIMA model that incorporates the highly seasonal behavior observed in the data. Results show that predictions with reasonable accuracy up to six months in advance are produced, showing the usefulness of these techniques to anticipate state-level criminal patterns and inform public policy.
In 1972, at a symposium celebrating the 70th birthday of Paul Dirac, John Wheeler proclaimed that "the framework falls down for everything that one has ever called a law of physics". Responsible for this "breakage [...] among the laws of physics" was the general theory of relativity, more specifically its prediction of massive stars gravitationally collapsing to "black holes", a term Wheeler himself had made popular some years earlier. In our paper, we investigate how Wheeler reached the conclusion that gravitational collapse calls into question the lawfulness of physics and how, subsequently, he tried to develop a new worldview, rethinking in his own way the lessons of quantum mechanics as well as drawing inspiration from other disciplines, not least biology.
Fran Casino, Claudia Pina, Pablo López-Aguilar
et al.
Digital evidence underpin the majority of crimes as their analysis is an integral part of almost every criminal investigation. Even if we temporarily disregard the numerous challenges in the collection and analysis of digital evidence, the exchange of the evidence among the different stakeholders has many thorny issues. Of specific interest are cross-border criminal investigations as the complexity is significantly high due to the heterogeneity of legal frameworks which beyond time bottlenecks can also become prohibiting. The aim of this article is to analyse the current state of practice of cross-border investigations considering the efficacy of current collaboration protocols along with the challenges and drawbacks to be overcome. Further to performing a legally-oriented research treatise, we recall all the challenges raised in the literature and discuss them from a more practical yet global perspective. Thus, this article paves the way to enabling practitioners and stakeholders to leverage horizontal strategies to fill in the identified gaps timely and accurately.
. Electronic evidence in the form of information and / or electronic documents is valid evidence in the Criminal Procedure Code that can substitute letters and serves as independent evidence and an extension of evidence. These three statuses are not mentioned the Criminal Procedure Code (KUHAP), but the regulations are explained in some special laws and legal instruments issued by the Supreme Court (MA). Electronic evidence is valid in criminal justice practice and is regulated in some special laws and legal instruments issued by the Supreme Court (MA). This normative legal research provides accurate description of the current existing facts in the form of secondary data on primary legal material or secondary materials for tertiary legal materials. It was found that the concept of audio-visual recordings as evidence is valid if the audio-visual recordings substantively describe the occurrence of a criminal event. The recognition of audio-visual recordings as evidence is based on the rights and obligations of the perpetrators and victims of criminal acts. Every perpetrator has the responsibility regarding general criminal acts that have been committed without being limited by inhibited by the general criminal procedure law. Every victim of a crime has the right to receive compensation and rehabilitation for the suffering and loss that they have experienced without being intervened by the general criminal procedure law. This new concept regards audio visual recording that describes the occurrence of a criminal act must be taken as evidence being verified for its authenticity. The verification can be carried out using a variety of methods aimed at assessing the truth and authenticity of the legal facts shown in the audio-visual recording.
Indonesia adheres to an integrated criminal justice system, which is the Criminal Procedure Code (KUHAP). There has been widespread criticism of law enforcement's reality in Indonesia, especially on the criminal justice system's performance, justice for law, and people's rights to legal certainty. Formulation of the research problem: How are the efforts to create a justice system of law enforcement in the criminal justice system? The research findings show that to create a justice system of law enforcement in the current Indonesian criminal justice system, besides referring to the crime control model and due process model, namely the moral of law enforcers to realize a sense of justice. Law, legal protection for both individuals and society, as well as legal certainty as norms agreed upon by organization in the form of national law which reflects the socio-cultural and religious values as manifestations and philosophical values of Pancasila as the source of Indonesian law and finally guarantees its maintenance in the criminal justice process.
To evaluate the promising advantages of massively parallel sequencing (MPS) in our casework, we analysed a total of 33 Y-chromosomal short tandem repeats (Y-STRs) with traditional capillary electrophoresis (CE) and 25 Y-STRs using the newer MPS technology. We studied the outcome of both technologies in 64 father-son pairs using stock and custom-designed kits. Current MPS technology confirmed the 13 mutational events observed with CE and improved our understanding of the complex nature of STR mutations. By detecting isometric sequence variants between unrelated males, we show that sequencing Y-STRs using MPS can boost discrimination power.
Criminal law and procedure, Public aspects of medicine
Early studies of risk assessment algorithms used in criminal justice revealed widespread racial biases. In response, machine learning researchers have developed methods for fairness, many of which rely on equalizing empirical metrics across protected attributes. Here, I recall sociotechnical perspectives to delineate the significant gap between fairness in theory and practice, focusing on criminal justice. I (1) illustrate how social context can undermine analyses that are restricted to an AI system's outputs, and (2) argue that much of the fair ML literature fails to account for epistemological issues with underlying crime data. Instead of building AI that reifies power imbalances, like risk assessment algorithms, I ask whether data science can be used to understand the root causes of structural marginalization.