Groups of firms often achieve a competitive advantage through the formation geo-industrial clusters. Although many exemplary clusters, such as Hollywood or Silicon Valley, have been frequently studied, systematic approaches to identify and analyze hierarchical structure clusters at global scale are rare. In this work, we use LinkedIn's employment histories more than 500 million users over 25 years construct labor flow network 4 across world apply recursive community detection algorithm...

Detecting violent behavior in videos to ensure public safety and security poses a significant challenge. Precisely identifying categorizing instances of violence real-life closed-circuit television, which vary across specifications locations, requires comprehensive understanding processing the sequential information embedded these videos. This study aims introduce model that adeptly grasps spatiotemporal context within diverse settings scenarios. We propose method accurately capture features...

Large language models (LLMs) with chat-based capabilities, such as ChatGPT, are widely used in various workflows. However, due to a limited understanding of these large-scale models, users struggle use this technology and experience different kinds dissatisfaction. Researchers have introduced several methods, prompt engineering, improve model responses. they focus on enhancing the model's performance specific tasks, little has been investigated how deal user dissatisfaction resulting from...

Artificial intelligence (AI) is becoming an indispensable part of both individual and organizational affairs, in terms rhetoric practices. However, many people who want to (and need to) work with AI may find the advent this technology overwhelming, as they are not well prepared do so. This paper aims address preparedness era by introducing two major agendas AI. Firstly, a diagnosis AI-related sentiments that inhibit working AI, four types "AI-phobia" presented: AI-literacy,...

Fashion is a multi-billion dollar industry with social and economic implications worldwide. To gain popularity, brands want to be represented by the top popular models. As new faces are selected using stringent (and often criticized) aesthetic criteria, priori predictions made difficult information cascades other fundamental trend-setting mechanisms. However, increasing usage of media within without may affecting this traditional system. We therefore seek understand ingredients success...

This paper examines network prominence in a co-prescription as an indicator of opioid doctor shopping (i.e., fraudulent solicitation opioids from multiple prescribers). Using longitudinal data large commercially insured population, we construct where tie between patients is weighted by the number shared prescribers. Given prior research suggesting that may be social process, hypothesize active shoppers will occupy central structural positions this network. We show prominence, operationalized...

In this paper, a design method based on genetic algorithm is proposed. Proposed useful for inexperienced engineers to select concept and it also offers optimized result. A voice-coil actuator optical image stabilization system in mobile phone camera designed by using method. To reduce the size, integrated with auto-focusing module Hall Effect sensor shares magnet voice-coil. Combined electromagnetic analysis used computing time. The final from fabricated verified experimentally. Experiment...

Abstract What determines the price of an artwork? This article leverages a comprehensive and novel dataset on art auctions contemporary artists to examine impact social visual features valuation artworks across global markets. Our findings indicate that signals allow us predict artwork exceptionally well, even approaching professionals’ prediction accuracy, while play marginal role. pattern is especially pronounced in emerging markets, supporting idea become more critical when it difficult...

In this paper, 2-axis voice coil actuator of optical image stabilization which is a technique to compensate trembling camera caused by vibration user's hand suggested. The designed for mobile phone camera. Considering total size and application auto focusing module, the concept moving auto-focusing module selected stabilization. handshake measured define specification actuator. Two kinds mechanisms made two prismatic joints are united electromagnetic part already designed. Finally, types...

The labor market is a complex ecosystem comprising diverse, interconnected entities, such as industries, occupations, skills, and firms.Due to the lack of systematic method map these heterogeneous entities together, each entity has been analyzed in isolation or only through pairwise relationships, inhibiting comprehensive understanding whole ecosystem.Here, we introduce Labor Space, vector-space embedding derived applying large language model with fine-tuning.Labor Space exposes relational...

질의-동영상 검색은 사용자가 입력한 질의와 가장 관련성 깊은 동영상을 찾는 분야이다. 이를 위해 기존 연구들은 적절히 잠재 벡터 공간으로 표상하여 검색에 활용한다. 하지만 동영상의 관련도를 계산하는 방식이 단순히 두 벡터의 내적값을 활용할 뿐, 의미, 내용, 설명가능성 등을 내포하지 않는다. 본 논문에서는 토큰 기반 공간에 위치한 임베딩으로 변환하여, 문서처럼 검색하여 의미적 유사도를 모델을 제안한다. 실험 결과 논문에서 제안하는 최종 모델의 성능은 베이스라인 대비 MSVD 데이터셋에서 Recall@1, Recall@5, Recall@10 기준 모두 향상된 성능을 보인다. 나아가 CLIP4Clip 제안 모델이 약 3.33배 빠르며, 최소한의 수정을 거친 BM25를 적용한 경우 208.11배 빠른 속도를 또한, 정성 평가를 통해 동영상으로부터 추출한 토큰이 자막에 준하는 보이며 구조임을 증명한다. Query-video retrieval is a field that finds...

We propose a method for extracting hierarchical backbones from bipartite network. Our leverages the observation that relationship between two nodes in network is often manifested as an asymmetry conditional probability of observing connections to them other node set. estimates both importance and direction pair nodes, thereby providing flexible way identify essential part networks. Using semi-synthetic benchmarks, we show our outperforms existing methods at identifying planted hierarchy...

This paper proposes an explainable machine learning model for consumer credit scoring in Mexico, emerging economy. We develop extreme gradient boosting (XGBoost) using non-traditional data from the Financial Inclusion National Survey. To address black box problem, we explore feature importance by estimating Shapley values that measure average marginal contributions across all possible subsets of features. The key drivers defaults include adverse economic effects due to COVID-19 pandemic and...

Large language models (LLMs) with chat-based capabilities, such as ChatGPT, are widely used in various workflows. However, due to a limited understanding of these large-scale models, users struggle use this technology and experience different kinds dissatisfaction. Researchers have introduced several methods, prompt engineering, improve model responses. they focus on enhancing the model's performance specific tasks, little has been investigated how deal user dissatisfaction resulting from...

This study investigates how the trade war between U.S. and China has influenced firm innovation. We examine impact of on patent applications filed by US firms that previously depended Chinese inventors. find decreased inventors who collaborated with them suffered a decline in patents when they relied substantially In response, these raised share engaged them. The findings suggest significance collaborative networks innovation we may create policy responses to future shocks.

This paper examines network prominence in a co-prescription as an indicator of doctor shopping (i.e., fraudulent solicitation prescriptions from multiple healthcare providers) for opioids. Using longitudinal data large commercially insured population, we construct where tie between patients was weighted by the number shared opioid prescribers. Given prior research suggesting that may be social process, hypothesize active shoppers will occupy central structural positions this network. We show...

Recently social network services(SNSs) have become important media to keep in touch with acquaintances as well share experiences of daily lives. Through these SNSs, users can directly photos, opinion and information. In this work, we implemented a web-based collaborative curation system which help generate album about specific common event. Our contains four parts: collecting contents from multiple web services, grouping those according sub-events, organizing the grouped addition explanatory...

In this paper we describe how to collect and group the contents of distributed heterogeneous SNSs for storytelling. particular, have addressed two issues storytelling, i.e. collecting grouping. At first, when users input their queries, our proposed system collects proper from using content-based filtering, followed by its grouping, based on similarity between content text information user query. Our implementation results show that a more informative detailed online story can be made SNS...