This paper addresses the computational scalability challenges in automating inspection tasks across large-scale spatial domains, where a fundamental trade-off exists between modeling fidelity for obstacle avoidance and planning tractability. While motivated by the autonomous inspection of offshore wind farms using Unmanned Surface Vessels (USVs), the proposed hierarchical framework offers a generalizable solution for navigating vast environments with sparse obstacles. Existing planners often suffer from prohibitive computational costs when processing environments with millions of grid cells. To overcome this, we present a novel multi-stage framework integrating hierarchical environment modeling with optimized search techniques. The core contributions are threefold: (1) A layered refinement grid strategy that unlike generic hierarchical structures, optimizes memory efficiency for extreme scale disparities to reduce the total grid count by 99.9%; (2) A hexagonal grid compression method that further decreases search nodes by over 95%, thereby mitigating directional bias and preserving topological clustering fidelity; and (3) A bidirectional shortcut-embedded A algorithm that executes efficient local detour planning within constrained corridors. Experimental validation in a large-scale real-world scenario (54 turbines) demonstrates the framework’s efficacy. The proposed method achieves hierarchically-refined path inspection paths in 1.21 seconds, significantly outperforming a baseline approach (5.97 seconds). This work provides a scalable, computationally efficient solution applicable to broad classes of automated inspection and logistics systems.

We develop efficient algorithms of exact simulation for quadratic stochastic intensity models that have become increasingly popular for modeling events arrivals, especially in economics, finance, and insurance. They have huge potential to be applied to many other areas such as operations management, queueing science, biostatistics, and epidemiology. Our algorithms are developed by the principle of exact distributional decomposition, which lies in a fully analytical expression for the joint Laplace transform of quadratic process and its integral newly derived in this paper. They do not involve any numerical Laplace inversion, have been validated by extensive numerical experiments, and substantially outperform all existing alternatives in the literature. Moreover, our algorithms are extendable to multidimensional point processes and beyond Cox processes to additionally incorporate two-sided random jumps with arbitrarily distributed sizes in the intensity for capturing self-exciting and self-correcting effects in event arrivals. Applications to portfolio loss modeling are provided to demonstrate the applicability and flexibility of our algorithms. History: Accepted by Bruno Tuffin, Area Editor for Simulation. Funding: This work was supported by the Beijing University of Posts and Telecommunications [Grant 2022RC58], the Shanghai University of Finance and Economics [Grant 2020110930], the National Natural Science Foundation of China [Grant 71401147], and Graduate Innovation Fund of Shanghai University of Finance and Economics [CXJJ-2023-387]. Supplemental Material: The software that supports the findings of this study is available within the paper and its Supplemental Information ( https://pubsonline.informs.org/doi/suppl/10.1287/ijoc.2023.0323 ) as well as from the IJOC GitHub software repository ( https://github.com/INFORMSJoC/2023.0323 ). The complete IJOC Software and Data Repository is available at https://informsjoc.github.io/ .

The switching capability required by data centers is rapidly developing, however, the technical limitations of the optical devices have created bottlenecks in their performance. Additionally, architectures that support thousands of servers remain to be explored. With a distributed control, we present a high-radix architecture that exploits the space domain and the wavelength domain to provide high throughput, achieving fast configuration and low latency. Considering the network size, we also propose two methods for implementation. We firstly form a strictly non-blocking network under the constraint of one connection per port. To scale up the switch, wavelength division multiplexing (WDM) is utilized to interconnect more servers and achieve higher bandwidth. The simulation results indicate that the proposed architecture can scale over thirty thousand servers with throughput reaching 87% while providing low latency.

Deep network compression has been achieved notable progress via knowledge distillation, where a teacher-student learning manner is adopted by using predetermined loss. Recently, more focuses have been transferred to employ the adversarial training to minimize the discrepancy between distributions of output from two networks. However, they always emphasize on result-oriented learning while neglecting the scheme of process-oriented learning, leading to the loss of rich information contained in the whole network pipeline. Whereas in other (non GAN-based) process-oriented methods, the knowledge have usually been transferred in a redundant manner. Observing that, the small network can not perfectly mimic a large one due to the huge gap of network scale, we propose a knowledge transfer method, involving effective intermediate supervision, under the adversarial training framework to learn the student network. Different from the other intermediate supervision methods, we design the knowledge representation in a compact form by introducing a task-driven attention mechanism. Meanwhile, to improve the representation capability of the attention-based method, a hierarchical structure is utilized so that powerful but highly squeezed knowledge is realized and the knowledge from teacher network could accommodate the size of student network. Extensive experimental results on three typical benchmark datasets, i.e., CIFAR-10, CIFAR-100, and ImageNet, demonstrate that our method achieves highly superior performances against state-of-the-art methods.

In this paper, we propose a new analytic unit for human mobility analysis -- the patch. We developed a process to identify Regularly Visited Patches (RVP) and a set of metrics to characterize and measure their spatial patterns. Using a large dataset of Foursquare check-ins as a test bed, we show that RVP analysis reveals fundamental patterns of human mobility and will lead to promising research with strong implications for businesses.

In this paper, we illustrate how User Generated Mobile Location Data (UGMLD) like Foursquare check-ins can be used in Trade Area Analysis (TAA) by introducing a new framework and corresponding analytic methods. Three key processes were created: identifying the activity center of a mobile user, profiling users based on their location history, and modeling users' preference probability. Extensions to traditional TAA are introduced, including customer-centric distance decay analysis and check-in sequence analysis. Adopting the rich content and context of UGMLD, these methods introduce new dimensions to modeling and delineating trade areas. Analyzing customers' visits to a business in the context of their daily life sheds new light on the nature and performance of the venue. This work has important business implications in the field of mobile computing.

In this work, we conducted a case study of a popular Chinese microblogging site, Sina-Weibo, to investigate how Chinese netizens used microblogging in response to a major disaster: the 2010 Yushu Earthquake. We combined multiple analysis methods in this case study, including content analysis of microblog messages, trend analysis of different topics, and an analysis of the information spreading process. This study helped us understand the roles played by microblogging systems in response to major disasters and enabled us to gain insight into how to harness the power of microblogging to facilitate disaster response. In addition, this work supplements existing works with an exploration of a non-Western socio-cultural system: how Chinese Internet users used microblogging in disaster response.

This is a case study about the early adoption and use of micro-blogging in a Fortune 500 company. The study used several independent data sources: five months of empirical micro-blogging data, user demographic information from corporate HR records, a web based survey, and targeted interviews. The results revealed that users vary in their posting activities, reading behaviors, and perceived benefits. The analysis also identified barriers to adoption, such as the noise-to-value ratio paradoxes. The findings can help both practitioners and scholars build an initial understanding of how knowledge workers are likely to use micro-blogging in the enterprise.

Pervasive and networked information and communication technology (ICT) has expanded the possibility for community participation in disaster response. This article describes the design and prototyping of a community-based emergency response system for a university campus. The main purposes of the system are to provide both top-down and bottom-up communication channels for information dissemination and gathering, and to facilitate peer-to-peer help within the community. Key design issues and challenges are discussed.

Making sense of research literature is a complicated process that involves various information seeking and compre-hension tasks. The lack of support for sensemaking in existing systems presents important design challenges and opportunities. This research proposes the design of an integral environment to support literature search, selection, organization and comprehension. Our system prototype, CiteSense, offers lightweight interaction tools and a smooth transition among various information activities. This research deepens our understanding of the design of systems that support the sensemaking of research literature.

Routing based on a connected dominating set (CDS) is a promising approach in wireless ad hoc networks. Wu and Li proposed a distributed approximation algorithm for calculating CDS in a given connected graph. However, this algorithm is difficult when there are obstacles in the network topology. Obstacle hybrid routing algorithm (OHRA), presented here, addresses these issues. OHRA consists of mobility model, CDS election and hybrid routing. In obstacle mobility model, we introduce STANDBY nodes as relaying nodes between two nodes that are invisible. This paper then proposes a distributed CDS election algorithm in the presence of obstacles. This algorithm extends Wu and Li's algorithm and utilizes STANDBY nodes to connect the existing dominating-nodes belonging to dominating set. In addition, OHRA based on a CDS uses hybrid routing scheme (flooding-based approach and position-based approach) to forward around any obstacles. Eventually, an example is given to show that the proposed approach can form a CDS and successfully construct routes.

Position based routing methods can construct effective routes; however, in some topologies with obstacles, such algorithms may end in failure. In order to avoid stationary obstacles, this paper presents a Voronoi-trajectory based hybrid routing (VTBR) algorithm that combines both proactive and reactive routing strategies to provide high routing performance. If the destination node is within the source node’s two-hop range, then the source will consult the routing table to determine the route. When the destination is beyond that range, the source establishes a trajectory which is the set of points of the shortest obstacle-avoiding path. The intermediate nodes take the forwarding decision based on the relationship to the trajectory. Eventually, an example and simulation experiments are conducted to validate and evaluate the performance of VTBR. Our simulation results show that the proposed algorithm makes it possible to successfully construct the routes.

At the NTCIR-4 workshop, Justsystem Corporation (JSC) and Clairvoyance Corporation (CC) collaborated in the cross-language retrieval task (CLIR). Our goal was to evaluate the performance and robustness of our recently developed commercial-grade CLIR systems for English and Asian languages. The main contribution of this article is the investigation of different strategies, their interactions in both monolingual and bilingual retrieval tasks, and their respective contributions to operational retrieval systems in the context of NTCIR-4. We report results of Japanese and English monolingual retrieval and results of Japanese-to-English bilingual retrieval. In monolingual retrieval analysis, we examine two special properties of the NTCIR experimental design (two levels of relevance and identical queries in multiple languages) and explore how they interact with strategies of our retrieval system, including pseudo-relevance feedback, multi-word term down-weighting, and term weight merging strategies. Our analysis shows that the choice of language (English or Japanese) does not have a significant impact on retrieval performance. Query expansion is slightly more effective with relaxed judgments than with rigid judgments. For better retrieval performance, weights of multi-word terms should be lowered. In the bilingual retrieval analysis, we aim to identify robust strategies that are effective when used alone and when used in combination with other strategies. We examine cross-lingual specific strategies such as translation disambiguation and translation structuring, as well as general strategies such as pseudo-relevance feedback and multi-word term down-weighting. For shorter title topics, pseudo-relevance feedback is a major performance enhancer, but translation structuring affects retrieval performance negatively when used alone or in combination with other strategies. All experimented strategies improve retrieval performance for the longer description topics, with pseudo-relevance feedback and translation structuring as the major contributors.

Multilingual applications frequently involve dealing with proper names, but names are often missing in bilingual lexicons.This problem is exacerbated for applications involving translation between Latin-scripted languages and Asian languages such as Chinese, Japanese and Korean (CJK) where simple string copying is not a solution.We present a novel approach for generating the ideographic representations of a CJK name written in a Latin script.The proposed approach involves first identifying the origin of the name, and then back-transliterating the name to all possible Chinese characters using language-specific mappings.To reduce the massive number of possibilities for computation, we apply a three-tier filtering process by filtering first through a set of attested bigrams, then through a set of attested terms, and lastly through the WWW for a final validation.We illustrate the approach with English-to-Japanese back-transliteration. Against test sets of Japanese given names and surnames, we have achieved average precisions of 73% and 90%, respectively.

The Web provides the largest, exploitable collection of language use. If we can mine the Web to build abstract models of language use, these models may have many applications. Here we present one example of using the implicit intelligence of language use to solve an important problem for machine translation programs and cross-lingual applications. This problem involves the translation of words written in katakana characters in Japanese. In this paper, we describe techniques of discovering katakana transliteration of English names and of finding English translations of multiword katakana sequences using implicit language models of English and Japanese found on the Web. These techniques were evaluated against human-constructed English-katakana glosses.

This paper introduces new interactive ways to create, manipulate and analyze shapes, even when those shapes do not have simple algebraic generators. This is made possible by using pixel-pattern rewrites to compute directly with bitmap representations. Such rewrites also permit the definition of functionality maps, bitmaps that specify the spatial scope of application functionality, and organic-widgets, implemented right in the pixels to have arbitrary form, integrated with the shape needs of the applications. Together these features should increase our capabilities for working with rich spatial domains.

For cross language information retrieval (CLIR) based on bilingual translation dictionaries, good performance depends upon lexical coverage in the dictionary. This is especially true for languages possessing few inter-language cognates, such as between Japanese and English. In this paper, we describe a method for automatically creating and validating candidate Japanese transliterated terms of English words. A phonetic English dictionary and a set of probabilistic mapping rules are used for automatically generating transliteration candidates. A monolingual Japanese corpus is then used for automatically validating the transliterated terms. We evaluate the usage of the extracted English-Japanese transliteration pairs with Japanese to English retrieval experiments over the CLEF bilingual test collections. The use of our automatically derived extension to a bilingual translation dictionary improves average precision, both before and after pseudo-relevance feedback, with gains ranging from 2.5% to 64.8%.

Attempts at discourse processing of spontaneously spoken dialogue face several difficulties: multiple hypotheses that result from the parser's attempts to make sense of the output from the speech recognizer, ambiguity that results from segmentation of multi-sentence utterances, and cumulative error -errors in the discourse context which cause further errors when subsequent sentences are processed.In this paper we will describe our robust parsers, our procedures for segmenting long utterances, and two approaches to discourse processing that attempt to deal with ambiguity and cumulative error.