Results 1 - 10
of
22
Event queries on correlated probabilistic streams (full version
, 2008
"... A major problem in detecting events in streams of data is that the data can be imprecise (e.g. RFID data). However, current state-of-the-art event detection systems such as Cayuga [14], SASE [46] or SnoopIB[1], assume the data is precise. Noise in the data can be captured using techniques such as hi ..."
Abstract
-
Cited by 61 (16 self)
- Add to MetaCart
(Show Context)
A major problem in detecting events in streams of data is that the data can be imprecise (e.g. RFID data). However, current state-of-the-art event detection systems such as Cayuga [14], SASE [46] or SnoopIB[1], assume the data is precise. Noise in the data can be captured using techniques such as hidden Markov models. Infer-ence on these models creates streams of probabilistic events which cannot be directly queried by existing systems. To address this challenge we propose Lahar1, an event processing system for prob-abilistic event streams. By exploiting the probabilistic nature of the data, Lahar yields a much higher recall and precision than deter-ministic techniques operating over only the most probable tuples. By using a novel static analysis and novel algorithms, Lahar pro-cesses data orders of magnitude more efficiently than a naïve ap-proach based on sampling. In this paper, we present Lahar’s static analysis and core algorithms. We demonstrate the quality and per-formance of our approach through experiments with our prototype implementation and comparisons with alternate methods.
Cascadia: a system for specifying, detecting, and managing RFID events
- In Proc. of the Sixth MobiSys Conf
, 2008
"... Cascadia is a system that provides RFID-based pervasive computing applications with an infrastructure for specifying, extracting and managing meaningful high-level events from raw RFID data. Cascadia provides three important services. First, it allows application developers and even users to specify ..."
Abstract
-
Cited by 20 (9 self)
- Add to MetaCart
(Show Context)
Cascadia is a system that provides RFID-based pervasive computing applications with an infrastructure for specifying, extracting and managing meaningful high-level events from raw RFID data. Cascadia provides three important services. First, it allows application developers and even users to specify events using either a declarative query language or an intuitive visual language based on direct manipulation. Second, it provides an API that facilitates the development of applications which rely on RFID-based events. Third, it automatically detects the specified events, forwards them to registered applications and stores them for later use (e.g., for historical queries). We present the design and implementation of Cascadia along with an evaluation that includes both a user study and measurements on traces collected in a building-wide RFID deployment. To demonstrate how Cascadia facilitates application development, we built a simple digital diary application in the form of a calendar that populates itself with RFID-based events. Cascadia copes with ambiguous RFID data and limitations in an RFID deployment by transforming RFID readings into probabilistic events. We show that this approach outperforms deterministic event detection techniques while avoiding the need to specify and train sophisticated models. 1.
Capturing data uncertainty in high-volume stream processing
- In CIDR
, 2009
"... We present the design and development of a data stream system that captures data uncertainty from data collection to query processing to final result generation. Our system focuses on data that is naturally modeled as continuous ran-dom variables such as many types of sensor data. To provide an end- ..."
Abstract
-
Cited by 14 (2 self)
- Add to MetaCart
(Show Context)
We present the design and development of a data stream system that captures data uncertainty from data collection to query processing to final result generation. Our system focuses on data that is naturally modeled as continuous ran-dom variables such as many types of sensor data. To provide an end-to-end solution, our system employs probabilistic modeling and inference to generate uncertainty description for raw data, and then a suite of statistical techniques to capture changes of uncertainty as data propagates through query operators. To cope with high-volume streams, we ex-plore advanced approximation techniques for both space and time efficiency. We are currently working with a group of scientists to evaluate our system using traces collected from real-world applications for hazardous weather monitoring and for object tracking and monitoring. 1.
Peex: Extracting probabilistic events from RFID data
, 2007
"... Radio-Frequency Identification (RFID) technology is increasingly being used to support various industrial and ubiquitous computing applications. Although this technology holds the promise to facilitate many of our every day activities, the noisy and low-level data produced by RFID readers today is e ..."
Abstract
-
Cited by 8 (3 self)
- Add to MetaCart
(Show Context)
Radio-Frequency Identification (RFID) technology is increasingly being used to support various industrial and ubiquitous computing applications. Although this technology holds the promise to facilitate many of our every day activities, the noisy and low-level data produced by RFID readers today is extremely difficult to use or comprehend in most but the simplest settings. In this paper, we present PEEX, a system that enables applications to easily define, extract, and manage meaningful probabilistic highlevel events from low-level RFID data. By using a declarative query language, the system simplifies definitions of new events. By using probabilities, the system copes with the noise and errors in the data and the inherent ambiguity in the event extraction. We have built PEEX as a layer on top of a traditional RDBMS, thus enabling applications not only to detect events but also manage them further as necessary. Through experiments with RFID traces collected on a real, building-wide RFID deployment, we demonstrate the performance and practicality of PEEX. 1.
Probabilistic rfid data management
, 2007
"... Radio Frequency Identification (RFID) technology is increasingly being used to improve various industrial processes, such as supply-chain management. Successes of this technology in industrial settings are leading many to consider other uses of RFID, including user-oriented public deployments. Howev ..."
Abstract
-
Cited by 6 (2 self)
- Add to MetaCart
(Show Context)
Radio Frequency Identification (RFID) technology is increasingly being used to improve various industrial processes, such as supply-chain management. Successes of this technology in industrial settings are leading many to consider other uses of RFID, including user-oriented public deployments. However, the noisy, low-level data produced by RFID readers is almost impossible to use or comprehend in most but the simplest settings. We present PEEX (Probabilistic Event EXtractor), a system that manages probabilistic high-level events from imprecise and erroneous RFID data. PEEX allows users to define probabilistic events from lower-level events. By using probabilities, the system copes with the noise in the data and the inherent ambiguity in the event extraction. We have built PEEX as a layer on top of a traditional RDBMS. We demonstrate, through experiments with real RFID traces collected on a small antenna deployment, that PEEX significantly improves event detection rates compared with deterministic techniques, and provides applications a flexible trade-off between event recall and precision. 1
Specification and Verification of Complex Location Events with Panoramic
"... Abstract. We present the design and evaluation of Panoramic, a tool that enables end-users to specify and verify an important family of complex location events. Our approach aims to reduce or eliminate critical barriers to deployment of emerging location-aware business activity monitoring applicatio ..."
Abstract
-
Cited by 4 (1 self)
- Add to MetaCart
(Show Context)
Abstract. We present the design and evaluation of Panoramic, a tool that enables end-users to specify and verify an important family of complex location events. Our approach aims to reduce or eliminate critical barriers to deployment of emerging location-aware business activity monitoring applications in domains like hospitals and office buildings. Panoramic does not require users to write code, understand complex models, perform elaborate demonstrations, generate test location traces, or blindly trust deterministic events. Instead, it allows end-users to specify and edit complex events with a visual language that embodies natural concepts of space and time. It also takes a novel approach to verification, in which events are extracted from historical sensor data traces and then presented with intelligible, hierarchical visualizations that represent uncertainty with probabilities. We build on our existing software for specifying and detecting events while enhancing it in non-trivial ways to facilitate event specification and verification. Our design is guided by a formative study with 12 non-programmers. We also use location traces from a building-scale radio frequency identification (RFID) deployment in a qualitative evaluation of Panoramic with 10 non-programmers. The results show that end-users can both understand and verify the behavior of complex location event specifications using Panoramic. 1
1 SPIRE: Efficient Data Inference and Compression over RFID Streams
"... challenges, including incomplete data, lack of location and containment information, and very high volumes. In this work, we present a novel data inference and compression substrate over RFID streams to address these challenges. Our substrate employs a time-varying graph model to efficiently capture ..."
Abstract
-
Cited by 3 (0 self)
- Add to MetaCart
(Show Context)
challenges, including incomplete data, lack of location and containment information, and very high volumes. In this work, we present a novel data inference and compression substrate over RFID streams to address these challenges. Our substrate employs a time-varying graph model to efficiently capture possible object locations and inter-object relationships such as containment from raw RFID streams. It then employs a probabilistic algorithm to estimate the most likely location and containment for each object. By performing such online inference, it enables online compression that recognizes and removes redundant information from the output stream of this substrate. We have implemented a prototype of our inference and compression substrate and evaluated it using both real traces from a laboratory warehouse setup and synthetic traces emulating enterprise supply chains. Results of a detailed performance study show that our data inference techniques provide high accuracy while retaining efficiency over RFID data streams, and our compression algorithm yields significant reduction in output data volume. Index Terms—RFID, data streams, data cleaning, compression, supply-chain management I.
Integrating Numbers and Words from the Ground Up
, 2006
"... All rights reserved. ..."
(Show Context)
