Large Language Models (LLM) have experienced strong development in recent years, with varied applications. This paper uses LLMs to develop a post-hoc process that provides more elaborated explanations of the results of food recommendation systems. By combining LLM with a hybrid extraction of key variables using SHAP, we obtain dynamic, convincing and more comprehensive explanations to lay user, compared to those in the literature. This approach enhances user trust and transparency by making complex recommendation outcomes easier to understand for a lay user.
Transformer-based models have revolutionized information retrieval, achieving state-of-the-art performance in document retrieval and ranking. For high-resource languages like English, an abundance of high-quality labeled datasets has facilitated the development of powerful models. However, developing powerful models for low-resource languages such as Arabic is challenging due to the scarcity of labeled data. While using translated English datasets can be considered to overcome the lack of labeled data, translated datasets have inherent information loss and inconsistencies introduced during the translation process. As a result, models fine-tuned on translated datasets typically underperform relative to their English counterparts. To address this issue, we explore the potential of transferring expertise from high-resource models to low-resource models. In particular, we investigate whether knowledge learned by English retrieval and reranking models can be effectively transferred to Arabic models via knowledge distillation. Our results demonstrate that knowledge distillation significantly improves the performance of Arabic information retrieval. Our models, fine-tuned using knowledge distillation on the mMARCO Arabic passage-ranking dataset, outperform state-of-the-art retrieval and reranker models. Specifically, our cross-encoder achieves an MRR@10 of 0.254, representing an 8% relative improvement over the previous best cross-encoder, mT5. In terms of recall, our bi-encoder achieves an R@1000 of 0.799, surpassing the late-interaction model mColBERT (R@1000 = 0.749, +6.7%) and the baseline BM25 (R@1000 = 0.637, +25%). Furthermore, by leveraging knowledge distillation with soft labels generated by an ensemble of IR models, we manage to achieve comparable or higher performance without requiring extensive manual annotation. This approach offers an effective mechanism for automatic annotation and pseudo-labeling in low-resource language scenarios.
Sharing Tacit Knowledge (TK) is a key objective of Knowledge Management, a critical issue for both academics and professional. In this work, we are interested in Software Tacit Knowledge, a new subfield of TK, which is very widespread, fairly frequent and remains unaddressed by existing TK approaches. We propose an approach to sharing Software TK between an expert and his learners to address a specific need. It is about the sharing of an expert’s TK relating to the mastery of a particular field, but also and especially relating to the control of the management of this field on dedicated software. This software continues to be widely used and even more so with telecommuting. The approach was developed in a conceptual framework and supported by a formal framework. The formal framework is cognitive to support human knowledge and formal to share it. The approach has been tested on a group of 97 first-year master’s students in a high school. Through the case study, the approach enabled our students to get the expected Software TK, i.e. the ability to design a conceptual solution (i.e. spreadsheet models in our case) for any design problem (i.e. linear problem in our case). Our approach database enables the extraction of key learning indicators that improve the sharing quality. The results are very encouraging. Our approach is applicable to any problem for any field and its software requiring both EK and TK. These softwares can run locally or online.
The issue of methodological support for the study of a police officer’s social portrait has been studied. The presented results of sociological research clearly reflect the dynamics of trust in the police and understanding of its role in society, as well as the police career attractiveness among the Russians. The author states that the methodological basis for a comprehensive study of a police officer’s social portrait is insufficiently developed. For a deeper and more complete study of a police officer’s social portrait, a generalized model has been proposed that includes the main structural elements. These include the research paradigm, object and subject, methods of data collection and processing, analysis, ethical principles, validity and reliability, as well as results interpretation. The formulated structural elements of the methodological research provide an integrated approach to the study of a police offi ’s social portrait. Paradigms for the study have been proposed: objectivist, based on quantitative methods, subjectivist, based on personal experience and qualitative methods, and mixed. The emphasis has been placed on the need to carefully choose a paradigm that has a key impact on the entire course of research activities. The present study is a valuable contribution to research methodology development in the law enforcement sphere, which makes it possible to optimize the processes of selecting, training, and managing law enforcement agencies personnel.
Deep neural networks (DNNs) are increasingly used to model brain computation, yet the principles linking their internal operations to neural mechanisms remain elusive. We propose a framework which leverages a recent advance in the interpretability of DNNs---mechanistic interpretability---to uncover shared algorithmic structure between DNNs and the brain via two parallel pathways. The Mechanism-to-Brain pathway uses interpretability to extract computational mechanisms from models that generate hypotheses about neural implementation. The Brain-to-Mechanism pathway begins from observed brain–model correspondences and applies interpretability to identify model components that could instantiate similar computations. These pathways interact through mutual constraints, ensuring that advances in one domain inform and delimit inquiry in the other, and through iterative refinement, where discrepancies between models and data drive the reciprocal revision of both mechanistic analyses and neuroscientific hypotheses. Together, they establish a principled route toward causal, mechanistic explanations of cognition, positioning DNNs as computational model organisms for probing the algorithms of the brain.
