44,721 Hits in 7.4 sec

Measuring system performance and topic discernment using generalized adaptive-weight mean

Chung Tong Lee, Vishwa Vinay, Eduarda Mendes Rodrigues, Gabriella Kazai, Nataša Milic-Frayling, Aleksandar Ignjatovic
2009 Proceeding of the 18th ACM conference on Information and knowledge management - CIKM '09  
We propose the Generalized Adaptive-Weight Mean (GAWM) measure and show how it can be computed as a fixed point of a function for which the Brouwer Fixed Point Theorem applies.  ...  Standard approaches to evaluating and comparing information retrieval systems compute simple averages of performance statistics across individual topics to measure the overall system performance.  ...  We develop the Generalized Adaptive-Weight Mean (GAWM) measure and present a unified model for system evaluation using GAWM, where the weights reflect the ability of the test topics to differentiate among  ... 
doi:10.1145/1645953.1646294 dblp:conf/cikm/LeeVRKMI09 fatcat:5b223b4mdrfoxi7arr2t6t62ie

Ad Hoc Monitoring of Vocabulary Shifts over Time

Tom Kenter, Melvin Wevers, Pim Huijnen, Maarten de Rijke
2015 Proceedings of the 24th ACM International on Conference on Information and Knowledge Management - CIKM '15  
We construct semantic networks of terms based on their representation in the semantic spaces and use graph-based measures to calculate saliency of terms.  ...  Word meanings change over time. Detecting shifts in meaning for particular words has been the focus of much research recently.  ...  A method relying on such terms would run a considerable risk of topic drift.We compute four measures of degree centrality: number of inlinks, weighted sum of inlinks, number of outlinks and weighted sum  ... 
doi:10.1145/2806416.2806474 dblp:conf/cikm/KenterWHR15 fatcat:yncfba6mujakje4sqyfsbjjg4i

A rough-fuzzy document grading system for customized text information retrieval

Shailendra Singh, Lipika Dey
2005 Information Processing & Management  
We have provided the detailed design of the various modules and some results related to the performance analysis of the system.  ...  In this paper, we have provided the design of a customized text information filtering system which learns user preferences and modifies the initial query to fetch better documents.  ...  Individual user ratings of pages are used to generate the user's profile adaptively, so that the recommendations gradually become more personalized.  ... 
doi:10.1016/j.ipm.2003.09.004 fatcat:lbyduyvjqzb5teu2ntqrzadgqy

Improving SVM Text Classification Performance through Threshold Adjustment [chapter]

James G. Shanahan, Norbert Roma
2003 Lecture Notes in Computer Science  
By using thresholds based on utility models and the ranking properties of classifiers, it is possible to overcome the precision bias of SVMs and insure robust performance in recall across a wide variety  ...  We describe an automatic process for adjusting the thresholds of generic SVM which incorporates a user utility model, an integral part of an information management system.  ...  Acknowledgements We thank David Evans, David Hull and Diane Kelly for their useful comments.  ... 
doi:10.1007/978-3-540-39857-8_33 fatcat:wxqw5xy46fejrbnnldxfiej2ii

Quality of health information on the Internet in pediatric neuro-oncology

Darren R. Hargrave, Ursula A. Hargrave, Eric Bouffet
2006 Neuro-Oncology  
The Web sites were evaluated for content quality by using the validated DISCERN rating instrument. Breadth of content and its accuracy were also scored by a checklist tool.  ...  Web sites were found deficient in topics covering etiology, late effects, prognosis, and treatment choices.  ...  The interobserver variability for both rating methods was measured by using either weighted or unweighted kappa statistics.  ... 
doi:10.1215/15228517-2005-008 pmid:16533758 pmcid:PMC1871939 fatcat:ts5gid2iurddrjd2calbygs4yu

Combining evidence for automatic Web session identification

Daqing He, Ayşe Göker, David J Harper
2002 Information Processing & Management  
Our previous work in traditional information retrieval systems has shown how using contextual information could improve retrieval performance.  ...  The approach automatically groups a user's consecutive search activities on the same search topic into one session.  ...  Bill Teahan and Dr. Andrei Petrovski for their comments.  ... 
doi:10.1016/s0306-4573(01)00060-7 fatcat:dwpixf7fprg7fp5eem3wrwynau

A semi-supervised document clustering technique for information organization

Han-Joon Kim, Sang-Goo Lee
2000 Proceedings of the ninth international conference on Information and knowledge management - CIKM '00  
Most clustering methods organizes documents into groups based only on similarity measures.  ...  As a basic clustering strategy, we use a variant of complete-linkage agglomerative hierarchical clustering, and develop the concepts (or seeds) of requested clusters by exploiting userrelevance feedback  ...  We use the standard measurements of information retrieval performance: precision, recall and F-measure for each cluster, with respect to the topic in question.  ... 
doi:10.1145/354756.354777 dblp:conf/cikm/KimL00 fatcat:54oxtfkqovh3zlzal7ubdz5lza

Networked Decentralized Control of Adaptive Structures

Alexander Warsewa, Institute for System Dynamics, University of Stuttgart, 70563 Stuttgart, Germany, Julia L. Wagner, Michael Böhm, Oliver Sawodny, Cristina Tarín
2020 Journal of Communications  
Control of such structures needs to be performed in real-time with high reliability and robustness. Decentralization provides a means to reduce cost and increase both redundancy and reliability.  ...  Subsystem models are derived from a global Finite Element (FE) model of the structure using model order reduction techniques which enables the use of distributed and Decentralized Information Filters (  ...  Information on the structural deformation is obtained using two different measurement systems -an optical camera system and strain gauge sensors.  ... 
doi:10.12720/jcm.15.6.496-502 fatcat:c47lkchkmrahhadxrmfwlu5jtm

Rough-fuzzy mlp: modular evolution, rule generation, and evaluation

S.K. Pal, S. Mitra, P. Mitra
2003 IEEE Transactions on Knowledge and Data Engineering  
The modular concept, based on "divide and conquer" strategy, provides accelerated training and a compact network suitable for generating a minimum number of rules with high certainty values.  ...  A methodology is described for evolving a Rough-fuzzy multi layer perceptron with modular concept using a genetic algorithm to obtain a structured network suitable for both classification and rule extraction  ...  ACKNOWLEDGMENTS This work was partly supported by a Council of Scientific and Industrial Research (CSIR) Grant 25(0039)/97/EMR-II, India.  ... 
doi:10.1109/tkde.2003.1161579 fatcat:t4famurdw5d3pbuz73v46x2j5y

An adaptive Web page recommendation service

Marko Balabanović
1997 Proceedings of the first international conference on Autonomous agents - AGENTS '97  
In this paper we explain the design of the system architecture and report the results of our rst experiment, evaluating recommendations provided to a group of test users.  ...  An adaptive recommendation service seeks to adapt to its users, providing increasingly personalized recommendations over time.  ...  Acknowledgments The research described was supported in part by the NSF/ARPA/NASA Digital Library project (NSF IRI-9411306) and in part by NSF contract IRI-9220645.  ... 
doi:10.1145/267658.267744 dblp:conf/agents/Balabanovic97 fatcat:qyjngf2qdjgw3k3ei7qcpxhgji

Improving M-learners' Performance Through Deep Learning Techniques by Leveraging Features Weights

Muhammad Adnan, Asad Habib, Jawad Ashraf, Babar Shah, Gohar Ali
2020 IEEE Access  
Subsequently, both models were selected for developing the M-learning model which included the performance categorization of M-learners under a five-level classification scheme and assigning weights to  ...  For an adaptive M-learning model, not only learning features are required, but it is also important to determine how they differ for various Mlearners, their weights, and interrelationship.  ...  Five-point Likert-scale was used to measure M-learners' satisfaction toward using the M-learning system where 5 means ''strongly agree'' and 1 means ''strongly disagree''.  ... 
doi:10.1109/access.2020.3007727 fatcat:ryls52oxwvgfvb6whjmg32z47a

AutoDiscern: rating the quality of online health information with hierarchical encoder attention-based neural networks

Laura Kinkead, Ahmed Allam, Michael Krauthammer
2020 BMC Medical Informatics and Decision Making  
We compared the performance of a traditional model (Random Forest) with that of a hierarchical encoder attention-based neural network (HEA) model using two language embeddings, BERT and BioBERT.  ...  We built an automated implementation of the DISCERN instrument (Brief version) using machine learning models.  ...  Funding The authors would like to acknowledge the Swiss National Science Foundation for their previous funding of A-Discern project (grant number P2TIP1-161635) awarded to AA. 1  ... 
doi:10.1186/s12911-020-01131-z pmid:32517759 fatcat:oje437dmbfc37nqddw6mgyndwi

Topic analysis and forecasting for science, technology and innovation: Methodology with a case study focusing on big data research

Yi Zhang, Guangquan Zhang, Hongshu Chen, Alan L. Porter, Donghua Zhu, Jie Lu
2016 Technological forecasting & social change  
By addressing and predicting these changes, this paper proposes an analytic method to (1) cluster associated terms and phrases to constitute meaningful technological topics and their interactions, and  ...  Our results are carried forward to present mechanisms that forecast prospective developments using Technology Roadmapping, combining qualitative and quantitative methodologies.  ...  Several aiding models are added as described below: 1) Cluster Validation Model Referring to the common performance measures in information retrieval, Recall, Precision, and F Measure are three target  ... 
doi:10.1016/j.techfore.2016.01.015 fatcat:zjqqq2pwondftgvl6zjbpq6c2y

Rapid speaker adaptation with speaker adaptive training and non-negative matrix factorization

Xueru Zhang, Kris Demuynck, Hugo Van hamme
2011 2011 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)  
The proposed fast adaptation algorithm, using a single enrollment sentence only, results in similar performance as fMLLR adapting on 40 enrollment sentences.  ...  In order to learn meaningful correlations between Gaussian activations from the train data, the NMF-based weight adaptation was combined with vocal tract length normalization (VTLN) and feature-space maximum  ...  The SD Gaussian mixture weights {λr} R r=1 are used to perform NMF based mixture weight speaker adaptation as described in the next section.  ... 
doi:10.1109/icassp.2011.5947343 dblp:conf/icassp/ZhangDh11 fatcat:cbxygadxwjezvkxr2gslrmt6u4

AutoDiscern: Rating the Quality of Online Health Information with Hierarchical Encoder Attention-based Neural Networks [article]

Laura Kinkead, Ahmed Allam, Michael Krauthammer
2020 arXiv   pre-print
We compared the performance of a traditional model (Random Forest) with that of a hierarchical encoder attention-based neural network (HEA) model using two language embeddings, BERT and BioBERT.  ...  We built an automated implementation of the DISCERN instrument (Brief version) using machine learning models.  ...  Acknowledgements The authors would like to acknowledge the Swiss National Science Foundation for their previous funding of A-Discern project (grant number P2TIP1-161635) awarded to AA.  ... 
arXiv:1912.12999v3 fatcat:ymrtvdlnrzdgdjmy7jr6bqliym
« Previous Showing results 1 — 15 out of 44,721 results