Interactions between humans and machines are often placed in a multi-layered network involving the multidimensional trust in communication, information, and socio-cognitive layers. In this complex environment, how to filter and fuse heterogeneous data is critical for effective decision making. In this work, we propose an ontology-based framework for information fusion, as a support system for human decision makers. In particular, we build upon the concept of composite trust, consisting of four trust types: communication trust, information trust, social trust, and cognitive trust. Based on the concept of multidimensional trust, we construct a composite trust ontology framework, called ComTrustO, that embraces four trust ontologies, one for each trust type. We present the details of the integrated ontology framework and discuss a concrete example scenario.
{"title":"ComTrustO: Composite trust-based ontology framework for information and decision fusion","authors":"A. Oltramari, Jin-Hee Cho","doi":"10.5281/ZENODO.32210","DOIUrl":"https://doi.org/10.5281/ZENODO.32210","url":null,"abstract":"Interactions between humans and machines are often placed in a multi-layered network involving the multidimensional trust in communication, information, and socio-cognitive layers. In this complex environment, how to filter and fuse heterogeneous data is critical for effective decision making. In this work, we propose an ontology-based framework for information fusion, as a support system for human decision makers. In particular, we build upon the concept of composite trust, consisting of four trust types: communication trust, information trust, social trust, and cognitive trust. Based on the concept of multidimensional trust, we construct a composite trust ontology framework, called ComTrustO, that embraces four trust ontologies, one for each trust type. We present the details of the integrated ontology framework and discuss a concrete example scenario.","PeriodicalId":297288,"journal":{"name":"2015 18th International Conference on Information Fusion (Fusion)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-07-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127149566","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
We compare several belief fusion methods, including the proportional conflict redistribution rules (PCR5 and PCR6) for multiple sources. The PCR fusion of evidence methods have shown improvement over the classical Dempster-Shafer and Bayesian fusion techniques in the presence of conflicting information. The PCR6 rule shows improvement over PCR5 when the number of sources increases. Using Hasse graphical diagrams, we highlight the comparison between the methods. To our knowledge, this is the first such comparison between PCR5 and PCR6 with more than two sources. The results point toward a transition between PCR5 and PCR6 at three sources.
{"title":"Information fusion with belief functions: A comparison of proportional conflict redistribution PCR5 and PCR6 rules for networked sensors","authors":"R. Ilin, Erik Blasch","doi":"10.5281/ZENODO.23211","DOIUrl":"https://doi.org/10.5281/ZENODO.23211","url":null,"abstract":"We compare several belief fusion methods, including the proportional conflict redistribution rules (PCR5 and PCR6) for multiple sources. The PCR fusion of evidence methods have shown improvement over the classical Dempster-Shafer and Bayesian fusion techniques in the presence of conflicting information. The PCR6 rule shows improvement over PCR5 when the number of sources increases. Using Hasse graphical diagrams, we highlight the comparison between the methods. To our knowledge, this is the first such comparison between PCR5 and PCR6 with more than two sources. The results point toward a transition between PCR5 and PCR6 at three sources.","PeriodicalId":297288,"journal":{"name":"2015 18th International Conference on Information Fusion (Fusion)","volume":"144 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-07-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131493894","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The increasing exploitation of the internet leads to new uncertainties, due to interdependencies and links between cyber and physical layers. As an example, the integration between telecommunication and physical processes, that happens when the power grid is managed and controlled, yields to epistemic uncertainty. Managing this uncertainty is possible using specific frameworks, usually coming from fuzzy theory such as Evidence Theory. This approach is attractive due to its flexibility in managing uncertainty by means of simple rule-based systems with data coming from heterogeneous sources. In this paper, Evidence Theory is applied in order to evaluate risk. Therefore, the authors propose a frame of discernment with a specific property among the elements based on a graph representation. This relationship leads to a smaller power set (called Reduced Power Set) that can be used as the classical power set, when the most common combination rules, such as Dempster or Smets, are applied. The paper demonstrates how the use of the Reduced Power Set yields to more efficient algorithms for combining evidences and to application of Evidence Theory for assessing risk.
由于网络和物理层之间的相互依赖和联系，对互联网的日益利用导致了新的不确定性。例如，当电网被管理和控制时，电信和物理过程之间的集成就会产生认知上的不确定性。管理这种不确定性可以使用特定的框架，通常来自模糊理论，如证据理论。这种方法很有吸引力，因为它在管理不确定性方面具有灵活性，可以通过简单的基于规则的系统来管理来自异构数据源的数据。本文运用证据理论对风险进行评价。因此，作者提出了一种基于图表示的元素间具有特定属性的识别框架。当应用最常见的组合规则(如Dempster或Smets)时，这种关系导致一个更小的功率集(称为Reduced power set)，它可以用作经典功率集。本文演示了如何使用约简功率集产生更有效的算法来组合证据和应用证据理论来评估风险。
{"title":"A graph-based evidence theory for assessing risk","authors":"Riccardo Santini, Chiara Foglietta, S. Panzieri","doi":"10.5281/ZENODO.32199","DOIUrl":"https://doi.org/10.5281/ZENODO.32199","url":null,"abstract":"The increasing exploitation of the internet leads to new uncertainties, due to interdependencies and links between cyber and physical layers. As an example, the integration between telecommunication and physical processes, that happens when the power grid is managed and controlled, yields to epistemic uncertainty. Managing this uncertainty is possible using specific frameworks, usually coming from fuzzy theory such as Evidence Theory. This approach is attractive due to its flexibility in managing uncertainty by means of simple rule-based systems with data coming from heterogeneous sources. In this paper, Evidence Theory is applied in order to evaluate risk. Therefore, the authors propose a frame of discernment with a specific property among the elements based on a graph representation. This relationship leads to a smaller power set (called Reduced Power Set) that can be used as the classical power set, when the most common combination rules, such as Dempster or Smets, are applied. The paper demonstrates how the use of the Reduced Power Set yields to more efficient algorithms for combining evidences and to application of Evidence Theory for assessing risk.","PeriodicalId":297288,"journal":{"name":"2015 18th International Conference on Information Fusion (Fusion)","volume":"101 14","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-07-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"120823386","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Zhunga Liu, Q. Pan, J. Dezert, Arnaud Martin, G. Mercier
The influence of the missing values in the classification of incomplete pattern mainly depends on the context. In this paper, we present a fast classification method for incomplete pattern based on the fusion of belief functions where the missing values are selectively (adaptively) estimated. At first, it is assumed that the missing information is not crucial for the classification, and the object (incomplete pattern) is classified based only on the available attribute values. However, if the object cannot be clearly classified, it implies that the missing values play an important role to obtain an accurate classification. In this case, the missing values will be imputed based on the K-nearest neighbor (K-NN) and self-organizing map (SOM) techniques, and the edited pattern with the imputation is then classified. The (original or edited) pattern is respectively classified according to each training class, and the classification results represented by basic belief assignments (BBA's) are fused with proper combination rules for making the credal classification. The object is allowed to belong with different masses of belief to the specific classes and meta-classes (i.e. disjunctions of several single classes). This credal classification captures well the uncertainty and imprecision of classification, and reduces effectively the rate of misclassifications thanks to the introduction of meta-classes. The effectiveness of the proposed method with respect to other classical methods is demonstrated based on several experiments using artificial and real data sets.
{"title":"Classification of incomplete patterns based on the fusion of belief functions","authors":"Zhunga Liu, Q. Pan, J. Dezert, Arnaud Martin, G. Mercier","doi":"10.5281/ZENODO.23204","DOIUrl":"https://doi.org/10.5281/ZENODO.23204","url":null,"abstract":"The influence of the missing values in the classification of incomplete pattern mainly depends on the context. In this paper, we present a fast classification method for incomplete pattern based on the fusion of belief functions where the missing values are selectively (adaptively) estimated. At first, it is assumed that the missing information is not crucial for the classification, and the object (incomplete pattern) is classified based only on the available attribute values. However, if the object cannot be clearly classified, it implies that the missing values play an important role to obtain an accurate classification. In this case, the missing values will be imputed based on the K-nearest neighbor (K-NN) and self-organizing map (SOM) techniques, and the edited pattern with the imputation is then classified. The (original or edited) pattern is respectively classified according to each training class, and the classification results represented by basic belief assignments (BBA's) are fused with proper combination rules for making the credal classification. The object is allowed to belong with different masses of belief to the specific classes and meta-classes (i.e. disjunctions of several single classes). This credal classification captures well the uncertainty and imprecision of classification, and reduces effectively the rate of misclassifications thanks to the introduction of meta-classes. The effectiveness of the proposed method with respect to other classical methods is demonstrated based on several experiments using artificial and real data sets.","PeriodicalId":297288,"journal":{"name":"2015 18th International Conference on Information Fusion (Fusion)","volume":"14 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-07-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132857039","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Grid map offers a useful representation of the perceived world for mobile robotics navigation. It will play a major role for the safety (obstacle avoidance) of next generations of terrestrial vehicles, as well as for future autonomous navigation systems. In a grid map, the occupancy state of each cell represents a small piece of information of the surrounding area of the vehicle. The state of each cell must be estimated from sensors measurements and classified in order to get a complete and precise perception of the dynamic environment where the vehicle moves. So far, the estimation and the grid map updating have been done using fusion techniques based on the probabilistic framework, or on the classical belief function framework thanks to an inverse model of the sensors and Dempster-Shafer rule of combination. Recently we have shown that PCR6 rule (Proportional Conflict Redistribution rule #6) proposed in DSmT (Dezert-Smarandache Theory) did improve substantially the quality of grid map with respect to other techniques, especially when the quality of available information is low, and when the sources of information appear as conflicting. In this paper, we go further and we analyze the performance of the improved version of PCR6 with Zhang's degree of intersection. We will show through different realistic scenarios (based on a LIDAR sensor) the benefit of using this new rule of combination in a practical application.
{"title":"Environment perception using grid occupancy estimation with belief functions","authors":"J. Dezert, J. Moras, B. Pannetier","doi":"10.5281/ZENODO.23208","DOIUrl":"https://doi.org/10.5281/ZENODO.23208","url":null,"abstract":"Grid map offers a useful representation of the perceived world for mobile robotics navigation. It will play a major role for the safety (obstacle avoidance) of next generations of terrestrial vehicles, as well as for future autonomous navigation systems. In a grid map, the occupancy state of each cell represents a small piece of information of the surrounding area of the vehicle. The state of each cell must be estimated from sensors measurements and classified in order to get a complete and precise perception of the dynamic environment where the vehicle moves. So far, the estimation and the grid map updating have been done using fusion techniques based on the probabilistic framework, or on the classical belief function framework thanks to an inverse model of the sensors and Dempster-Shafer rule of combination. Recently we have shown that PCR6 rule (Proportional Conflict Redistribution rule #6) proposed in DSmT (Dezert-Smarandache Theory) did improve substantially the quality of grid map with respect to other techniques, especially when the quality of available information is low, and when the sources of information appear as conflicting. In this paper, we go further and we analyze the performance of the improved version of PCR6 with Zhang's degree of intersection. We will show through different realistic scenarios (based on a LIDAR sensor) the benefit of using this new rule of combination in a practical application.","PeriodicalId":297288,"journal":{"name":"2015 18th International Conference on Information Fusion (Fusion)","volume":"161 12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-07-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129048692","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This paper proposes a new generic object recognition (GOR) method based on the multiple feature fusion of 2D and 3D SIFT (scale invariant feature transform) descriptors drawn from 2D images and 3D point clouds. We also use trained Support Vector Machine (SVM) classifiers to recognize the objects from the result of the multiple feature fusion. We analyze and evaluate different strategies for making this multiple feature fusion applied to real open-datasets. Our results show that this new GOR method has higher recognition rates than classical methods, even if one has large intra-class variations, or high inter-class similarities of the objects to recognize, which demonstrates the potential interest of this new approach.
{"title":"Generic object recognition based on the fusion of 2D and 3D SIFT descriptors","authors":"Miaomiao Liu, Xinde Li, J. Dezert, C. Luo","doi":"10.5281/ZENODO.23200","DOIUrl":"https://doi.org/10.5281/ZENODO.23200","url":null,"abstract":"This paper proposes a new generic object recognition (GOR) method based on the multiple feature fusion of 2D and 3D SIFT (scale invariant feature transform) descriptors drawn from 2D images and 3D point clouds. We also use trained Support Vector Machine (SVM) classifiers to recognize the objects from the result of the multiple feature fusion. We analyze and evaluate different strategies for making this multiple feature fusion applied to real open-datasets. Our results show that this new GOR method has higher recognition rates than classical methods, even if one has large intra-class variations, or high inter-class similarities of the objects to recognize, which demonstrates the potential interest of this new approach.","PeriodicalId":297288,"journal":{"name":"2015 18th International Conference on Information Fusion (Fusion)","volume":"57 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-07-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124628982","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
We develop a novel information fusion scheme based on topological event space, viewed as a distributive lattice. We discuss the advantages of topological modeling and compare our approach to the existing Bayesian, Dempster-Shafer, and Dezert-Smarandache approaches. The proposed scheme is described in detail and illustrated with an example of fusion of three sensors in the presence of missing information.
{"title":"Information fusion with topological event spaces","authors":"R. Ilin, Jun Zhang","doi":"10.5281/ZENODO.23214","DOIUrl":"https://doi.org/10.5281/ZENODO.23214","url":null,"abstract":"We develop a novel information fusion scheme based on topological event space, viewed as a distributive lattice. We discuss the advantages of topological modeling and compare our approach to the existing Bayesian, Dempster-Shafer, and Dezert-Smarandache approaches. The proposed scheme is described in detail and illustrated with an example of fusion of three sensors in the presence of missing information.","PeriodicalId":297288,"journal":{"name":"2015 18th International Conference on Information Fusion (Fusion)","volume":"16 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-07-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116198204","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In this paper, we present a method to estimate the quality (trustfulness) of the solutions of the classical optimal data association (DA) problem associated with a given source of information (also called a criterion). We also present a method to solve the multi-criteria DA problem and to estimate the quality of its solution. Our approach is new and mixes classical algorithms (typically Murty's approach coupled with Auction) for the search of the best and the second best DA solutions, and belief functions (BF) with PCR6 (Proportional Conflict Redistribution rule # 6) combination rule drawn from DSmT (Dezert-Smarandache Theory) to establish the quality matrix of the global optimal DA solution. In order to take into account the importances of criteria in the fusion process, we use weighting factors which can be derived by different manners (ad-hoc choice, quality of each local DA solution, or inspired by Saaty's Analytic Hierarchy Process (AHP)). A simple complete example is provided to show how our method works and for helping the reader to verify by him or herself the validity of our results.
{"title":"On the quality estimation of optimal multiple criteria data association solutions","authors":"J. Dezert, K. Benameur, L. Ratton, J. Grandin","doi":"10.5281/ZENODO.23202","DOIUrl":"https://doi.org/10.5281/ZENODO.23202","url":null,"abstract":"In this paper, we present a method to estimate the quality (trustfulness) of the solutions of the classical optimal data association (DA) problem associated with a given source of information (also called a criterion). We also present a method to solve the multi-criteria DA problem and to estimate the quality of its solution. Our approach is new and mixes classical algorithms (typically Murty's approach coupled with Auction) for the search of the best and the second best DA solutions, and belief functions (BF) with PCR6 (Proportional Conflict Redistribution rule # 6) combination rule drawn from DSmT (Dezert-Smarandache Theory) to establish the quality matrix of the global optimal DA solution. In order to take into account the importances of criteria in the fusion process, we use weighting factors which can be derived by different manners (ad-hoc choice, quality of each local DA solution, or inspired by Saaty's Analytic Hierarchy Process (AHP)). A simple complete example is provided to show how our method works and for helping the reader to verify by him or herself the validity of our results.","PeriodicalId":297288,"journal":{"name":"2015 18th International Conference on Information Fusion (Fusion)","volume":"67 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-07-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122354833","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The theory of belief functions is a very appealing theory for uncertainty modeling and reasoning which has been widely used in information fusion. However, when the cardinality of the frame of discernment and the number of the focal elements are large the fusion of belief functions requires in general a high computational complexity. To circumvent this difficulty, many methods were proposed to implement more efficiently the combination rules and to approximate basic belief assignments (BBA's) into simplest ones to reduce the number of focal elements involved in the fusion process. In this paper, we present a novel principle for approximating a BBA by withdrawing more redundant focal elements of the original BBA. Two methods based on this principle are presented (using batch and recursive implementations). Numerical examples, simulations and related analyses are provided to illustrate and evaluate the performances of this new BBA approximation method.
{"title":"Two novel methods for BBA approximation based on focal element redundancy","authors":"Deqiang Han, J. Dezert, Yi Yang","doi":"10.5281/ZENODO.23206","DOIUrl":"https://doi.org/10.5281/ZENODO.23206","url":null,"abstract":"The theory of belief functions is a very appealing theory for uncertainty modeling and reasoning which has been widely used in information fusion. However, when the cardinality of the frame of discernment and the number of the focal elements are large the fusion of belief functions requires in general a high computational complexity. To circumvent this difficulty, many methods were proposed to implement more efficiently the combination rules and to approximate basic belief assignments (BBA's) into simplest ones to reduce the number of focal elements involved in the fusion process. In this paper, we present a novel principle for approximating a BBA by withdrawing more redundant focal elements of the original BBA. Two methods based on this principle are presented (using batch and recursive implementations). Numerical examples, simulations and related analyses are provided to illustrate and evaluate the performances of this new BBA approximation method.","PeriodicalId":297288,"journal":{"name":"2015 18th International Conference on Information Fusion (Fusion)","volume":"43 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-07-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126811401","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In this paper, we propose a modification of PCR5 and PCR6 fusion rules with degrees of intersections for taking into account the cardinality of focal elements of each source of evidence to combine. We show in very simple examples the interest of these new fusion rules w.r.t. classical Dempster-Shafer, PCR6, Zhang's and Jaccard's Center rules of combination.
{"title":"Modified PCR rules of combination with degrees of intersections","authors":"F. Smarandache, J. Dezert","doi":"10.5281/ZENODO.48919","DOIUrl":"https://doi.org/10.5281/ZENODO.48919","url":null,"abstract":"In this paper, we propose a modification of PCR5 and PCR6 fusion rules with degrees of intersections for taking into account the cardinality of focal elements of each source of evidence to combine. We show in very simple examples the interest of these new fusion rules w.r.t. classical Dempster-Shafer, PCR6, Zhang's and Jaccard's Center rules of combination.","PeriodicalId":297288,"journal":{"name":"2015 18th International Conference on Information Fusion (Fusion)","volume":"138 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-07-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131948303","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：