Search

Accurate PM2.5 prediction is part of the fight against air pollution that helps governments to manage environmental policy. Satellite Remote sensing aerosol optical depth (AOD) processed by The Multi-Angle Implementation of Atmospheric Correlation (MAIAC) algorithm allows us to observe the transportation of remote pollutants between regions. The paper proposes a composite neural network model, the Remote Transported Pollutants (RTP) model, for such long-range pollutant transportation that predicts more accurate local PM2.5 concentrations given such satellite data. The proposed RTP model integrates several deep learning components and learns from the heterogeneous features of various domains. We also detected remote transportation pollution events (RTPEs) at two reference sites from the AOD data. Extensive experiments using real-world data show that the proposed RTP model outperforms the base model that does not account for RTPEs by 17%-30%, 23%-26% and 18%-22% and state-of-the-art models that account for RTPEs by 12%-22%, 12%-14%, and 10%-11% at +4h to +24h, +28h to +48 hours, and +52h to +72h hours respectively. [ABSTRACT FROM AUTHOR]

When an event occurs, it attracts attention of information sources to publish related documents along its lifespan. The task of event detection is to automatically identify events and their related documents from a document stream, which is a set of chronologically ordered documents collected from various information sources. Generally, each event has a distinct activeness development so that its status changes continuously during its lifespan. When an event is active, there are a lot of related documents from various information sources. In contrast when it is inactive, there are very few documents, but they are focused. Previous works on event detection did not consider the characteristics of the event's activeness, and used rigid thresholds for event detection. We propose a concept called life profile, modeled by a hidden Markov model, to model the activeness trends of events. In addition, a general event detection framework, LIPED, which utilizes the learned life profiles and the burst-and-diverse characteristic to adjust the event detection thresholds adaptively, can be incorporated into existing event detection methods. Based on the official TDT corpus and contest rules, the evaluation results show that existing detection methods that incorporate LIPED achieve better performance in the cost and F1 metrics, than without. Categories and Subject Descriptors: H.3.3 [Information Storage and Retrieval]: Information Search and Retrieval--Clustering; H2.8 [Database Management]: Database Applications--Data mining; G.3 [Probability and Statistics]: Time series analysis General Terms: Algorithm, Design, Experimentation Additional Key Words and Phrases: Event detection, topic detection, TDT, life profiles, hidden Markov models, clustering

To link to full-text access for this article, visit this link: http://dx.doi.org/10.1016/j.comnet.2007.11.020 Byline: Yung-Cheng Tu (a), Meng Chang Chen (b), Yeali S. Sun (c), Wei-Kuan Shih (a) Keywords: Bandwidth allocation; End-to-end delay; Bulk scheduling; Quality of service Abstract: Providing end-to-end delay guarantees for delay sensitive applications is an important packet scheduling issue with routers. In this paper, to support end-to-end delay requirements, we propose a novel network scheduling scheme, called the bulk scheduling scheme (BSS), which is built on top of existing schedulers of intermediate nodes without modifying transmission protocols on either the sender or receiver sides. By inserting special control packets, which called TED (Traffic Specification with End-to-end Deadline) packets, into packet flows at the ingress router periodically, the BSS schedulers of the intermediate nodes can dynamically allocate the necessary bandwidth to each flow to enforce the end-to-end delay, according to the information in the TED packets. The introduction of TED packets incurs less overhead than the per-packet marking approaches. Three flow bandwidth estimation methods are presented, and their performance properties are analyzed. BSS also provides a dropping policy for discarding late packets and a feedback mechanism for discovering and resolving bottlenecks. The simulation results show that BSS performs efficiently as expected. Author Affiliation: (a) Department of Computer Science, National Tsing Hua University, Hsinchu, Taiwan, ROC (b) Institute of Information Science, Academia Sinica, Taipei, Taiwan, ROC (c) Department of Information Management, National Taiwan University, Taipei, Taiwan, ROC Article History: Received 21 February 2007; Revised 4 November 2007; Accepted 27 November 2007 Article Note: (miscellaneous) Responsible Editor: P. Frossard

The goal of online event analysis is to detect events and track their associated documents in real time from a continuous stream of documents generated by multiple information sources. Unlike traditional text categorization methods, event analysis approaches consider the temporal relations among documents. However, such methods suffer from the threshord-dependency problem, so they only perform well for a narrow range of thresholds. In addition, if the contents of a document stream change, the optimal threshold (that is, the threshold that yields the best performance) often changes as well. In this paper, we propose a thresholdresilient online algorithm, called the Incremental Probabilistic Latent Semantic Indexing (IPLSI) algorithm, which alleviates the threshold-dependency problem and simultaneously maintains the continuity of the latent semantics to better capture the story fine development of events. The IPLSI algorithm is theoretically sound and empirically efficient and effective for event analysis. The results of the performance evaluation performed on the Topic Detection and Tracking (TDT)-4 corpus show that the algorithm reduces the cost of event analysis by as much as 15 percent ~ 20 percent and increases the acceptable threshold range by 200 percent to 300 percent over the baseline. Index Terms--Clustering, online event analysis, probabilistic algorithms, text mining.

In this paper, we propose an interframe-space (IFS)-based distributed-fair-queuing (IDFQ) mechanism to provide proportional fairness service for IEEE 802.11 wireless local-area networks (WLANs). IDFQ is designed to emulate self-clocked fair queuing in a distributed manner. It eliminates the backoff process as implemented in existing work and introduces a new mechanism to assign the IFS value to each station. IDFQ is immune from the implementation problem suffered by existing IFS-based mechanisms and is adaptive to the collision state in the system. Moreover, it can be used to eliminate the performance-anomaly problem of 802.11 medium-access control. The performance of IDFQ is validated by ns-2 simulations. The simulation results show that IDFQ supports fairness service for flows in proportion to their weights and outperforms existing mechanisms in terms of fairness and stability, rendering IDFQ an excellent candidate to provide weighted fairness in IEEE 802.11 WLANs. Index Terms--Distributed fair queuing (DFS), IEEE 802.11 wireless local-area networks (WLANs).

To link to full-text access for this article, visit this link: http://dx.doi.org/10.1016/j.comnet.2006.07.013 Byline: Jeng Farn Lee (a)(c), Meng Chang Chen (a), Yeali Sun (b) Keywords: Scheduling; Rate control; Weighted fair; QoS Abstract: While existing weighted fair scheduling schemes guarantee minimum bandwidths/resources for the classes/processes of a shared channel, the maximum rate control, which is critical to service providers, carriers, and network managers for resource management and business strategies in many applications, is generally enforced by employing traffic policing mechanisms. These approaches use either a concatenation of the rate controller and scheduler, or a policer in front of the scheduler. The concatenation method uses two sets of queues and a management apparatus that incurs overhead. The latter method may allow bursty traffic to pass through the controller, which violates the maximum rate constraint, or results in packet loss. In this paper, we present a new weighted fair scheduling scheme, WF.sup.2Q-M, which can simultaneously support maximum rate control and minimum service rate guarantees. WF.sup.2Q-M uses the virtual clock adjustment method to enforce maximum rate control and distribute the excess bandwidths of saturated sessions to other sessions without recalculating the virtual starting and finishing times of sessions. In terms of performance, we prove that WF.sup.2Q-M is theoretically bounded by a corresponding fluid reference model. A procedural scheduling implementation of WF.sup.2Q-M is proposed, and a proof of correctness is given. Finally, we present the results of extensive experiments to show that the performance of WF.sup.2Q-M is just as claimed. Author Affiliation: (a) Institute of Information Science, Academia Sinica Taiwan, Taipei 115, Taiwan, ROC (b) Department of Information Management, National Taiwan University, Taiwan, ROC (c) Department of Electrical Engineering, National Taiwan University, Taipei 106, Taiwan, ROC Article History: Received 3 November 2004; Revised 4 May 2006; Accepted 27 July 2006 Article Note: (miscellaneous) Responsible Editor: R. Sivakumar

An event can be described by a sequence of chronological documents from several information sources that together describe a story or happening. The goal of event detection and tracking is to automatically identify events and their associated documents during their life cycles. Conventional document clustering and classification techniques cannot effectively detect and track sequential events, as they ignore the temporal relationships among documents related to an event. The life cycle of an event is analogous to living beings. With abundant nourishment (i.e., related documents for the event), the life cycle is prolonged; conversely, an event or living fades away when nourishment is exhausted. Improper tracking algorithms often unnecessarily prolong or shorten the life cycle of detected events. In this paper, we propose an aging theory to model the life cycle of sequential events, which incorporates a traditional single-pass clustering algorithm to detect and track events. Our experiment results show that the proposed method achieves a better overall performance for both long-running and short-term events than previous approaches. Moreover, we find that the aging parameters of the aging schemes are profile dependent and that using proper profile-specific aging parameters improves the detection and tracking performance further. Index Terms--Clustering, knowledge life cycle, web mining.

To link to full-text access for this article, visit this link: http://dx.doi.org/10.1016/j.comnet.2006.06.005 Byline: Meng Chang Chen (a), Li-Ping Tung (b), Yeali S. Sun (c), Wei-Kuan Shih (b) Keywords: Differentiated services; Quality of Service; VoIP; Wireless LAN; IEEE 802.11 Abstract: In this paper, we propose a new per-CLAss Flow fixed proportional differentiated service model (CLAF) and a companioning medium access control scheme for multi-service wireless LANs (WLANs), based on the IEEE 802.11 framework. Different from conventional relative differentiated service, in CLAF, a fixed bandwidth quality ratio is guaranteed on per-class per-flow basis regardless of the traffic load of each service class. Specifically, each service class is assigned a number of separate coordination periods, proportional to the policy-based bandwidth quality ratio for class isolation. Each class is associated with its own contention window size which is dynamically adjusted in accordance with the number of flows in the class in such a way to minimize collision probability between flows of the same class. Simulations results of the CLAF performance as well as a comparison with IEEE 802.11e EDCA including the support of QoS-sensitive VoIP applications are presented. The results show that the proposed scheme outperforms EDCA and achieves better resource utilization efficiency. CLAF can provide users a more predictive, affirmative service guarantees than conventional relative differentiated service like IEEE 802.11e EDCA. Author Affiliation: (a) Institute of Information Science, Academia Sinica, NanKang, Taipei 115, Taiwan, ROC (b) Department of Computer Science, National Tsing Hua University, Hsinchu, Taiwan, ROC (c) Department of Information Management, National Taiwan University, Taipei, Taiwan, ROC Article History: Received 3 May 2006; Revised 13 June 2006; Accepted 18 June 2006 Article Note: (miscellaneous) Responsible Editor: X.S. Shen

This paper presents an intelligent Internet information system, Automatic Classifier for the Internet Resource Discovery (ACIRD), which uses machine learning techniques to organize and retrieve Internet documents. ACIRD consists of a knowledge acquisition process, document classifier, and two-phase search engine. The knowledge acquisition process of ACIRD automatically learns classification knowledge from classified Internet documents. The document classifier applies learned classification knowledge to classify newly collected Internet documents into one or more classes. Experimental results indicate that ACIRD performs as well or better than human experts in both knowledge acquisition and document classification. By using the learned classification knowledge and the given class lattice, the ACIRD two-phase search engine responds to user queries with hierarchically structured navigable results (instead of a conventional flat ranked document list), which greatly aids users in locating information from numerous, diversified Internet documents. Index Terms--Document classification, data mining, information retrieval, search engine.