Automatic image annotation methods require a quality training image dataset, from which annotations for target images are obtained. At present, the main problem with these methods is their low effectiveness and scalability if a large-scale training dataset is used. Current methods use only global image features for search. We proposed a method to obtain annotations for target images, which is based on a novel combination of local and global features during search stage. We are able to ensure the robustness and generalization needed by complex queries and significantly eliminate irrelevant results. In our method, in analogy with text documents, the global features represent words extracted from paragraphs of a document with the highest frequency of occurrence and the local features represent key words extracted from the entire document. We are able to identify objects directly in target images and for each obtained annotation we estimate the probability of its relevance. During search, we retrieve similar images containing the correct keywords for a given target image. For example, we prioritize images where extracted objects of interest from the target images are dominant as it is more likely that words associated with the images describe the objects. We tailored our method to use large-scale image training datasets and evaluated it with the Corel5K corpus which consists of 5000 images from 50 Corel Stock Photo CDs.
{"title":"Automatic Image Annotation Using Global and Local Features","authors":"M. Bieliková, Eduard Kuric","doi":"10.1109/SMAP.2011.14","DOIUrl":"https://doi.org/10.1109/SMAP.2011.14","url":null,"abstract":"Automatic image annotation methods require a quality training image dataset, from which annotations for target images are obtained. At present, the main problem with these methods is their low effectiveness and scalability if a large-scale training dataset is used. Current methods use only global image features for search. We proposed a method to obtain annotations for target images, which is based on a novel combination of local and global features during search stage. We are able to ensure the robustness and generalization needed by complex queries and significantly eliminate irrelevant results. In our method, in analogy with text documents, the global features represent words extracted from paragraphs of a document with the highest frequency of occurrence and the local features represent key words extracted from the entire document. We are able to identify objects directly in target images and for each obtained annotation we estimate the probability of its relevance. During search, we retrieve similar images containing the correct keywords for a given target image. For example, we prioritize images where extracted objects of interest from the target images are dominant as it is more likely that words associated with the images describe the objects. We tailored our method to use large-scale image training datasets and evaluated it with the Corel5K corpus which consists of 5000 images from 50 Corel Stock Photo CDs.","PeriodicalId":346975,"journal":{"name":"2011 Sixth International Workshop on Semantic Media Adaptation and Personalization","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2011-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115512560","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}
Our research is focused on semiautomatic domain model building from texts written in a natural language. The paper deals with defining the scope of the work and describes approaches of domain model creation and its representation. We describe the context of natural language processing focused on selected frequently solved tasks in this discipline and list some freely available tools suitable for text processing at the semantic level. We introduce appropriate tools and design methods to achieve determined goals. At the end of the paper, an experiment is conducted on which the successes of the implemented methods is verified, and suggestions and considerations for further development are presented.
{"title":"Semiautomatic Domain Model Building from Text-Data","authors":"Petr Šaloun, Zdenek Velart, Petr Klimanek","doi":"10.1109/SMAP.2011.18","DOIUrl":"https://doi.org/10.1109/SMAP.2011.18","url":null,"abstract":"Our research is focused on semiautomatic domain model building from texts written in a natural language. The paper deals with defining the scope of the work and describes approaches of domain model creation and its representation. We describe the context of natural language processing focused on selected frequently solved tasks in this discipline and list some freely available tools suitable for text processing at the semantic level. We introduce appropriate tools and design methods to achieve determined goals. At the end of the paper, an experiment is conducted on which the successes of the implemented methods is verified, and suggestions and considerations for further development are presented.","PeriodicalId":346975,"journal":{"name":"2011 Sixth International Workshop on Semantic Media Adaptation and Personalization","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2011-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130040789","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}
Nikolaos Tarantilis, Chrisa Tsinaraki, F. Kazasis, N. Gioldasis, S. Christodoulakis
In this paper we present EVISUGE, a system that allows the visualization and management of real-world scenarios on Google Earth. An EVISUGE scenario is composed of events, which are represented according to the MOME (Mobile Multimedia Event Capturing and Visualization) event representation model that we have developed. The scenario events are visualized on top of the Google Earth 3D interactive maps, with respect to their spatial and temporal features. We demonstrate the EVISUGE system through a real-world scenario: The specification and visualization of a naturalistic route.
{"title":"EVISUGE: Event Visualization on Google Earth","authors":"Nikolaos Tarantilis, Chrisa Tsinaraki, F. Kazasis, N. Gioldasis, S. Christodoulakis","doi":"10.1109/SMAP.2011.23","DOIUrl":"https://doi.org/10.1109/SMAP.2011.23","url":null,"abstract":"In this paper we present EVISUGE, a system that allows the visualization and management of real-world scenarios on Google Earth. An EVISUGE scenario is composed of events, which are represented according to the MOME (Mobile Multimedia Event Capturing and Visualization) event representation model that we have developed. The scenario events are visualized on top of the Google Earth 3D interactive maps, with respect to their spatial and temporal features. We demonstrate the EVISUGE system through a real-world scenario: The specification and visualization of a naturalistic route.","PeriodicalId":346975,"journal":{"name":"2011 Sixth International Workshop on Semantic Media Adaptation and Personalization","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2011-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131182553","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}
Zenonas Theodosiou, Olga Georgiou, N. Tsapatsoulis
Automatic semantic tagging of multimedia is still inefficient due to the difficulties in modelling abstract or complex terms using low level features. The degree of consensus and homogeneity in judgements among annotators is very important in semantic image and video retrieval. In this paper we present a novel method in evaluating the annotators consistency, which uses an innovative database schema and combines two different annotation approaches. A set of 100 images were annotated by 16 annotators using vocabulary keywords and free keywords. The results indicate that combination of annotation methods may lead to increased annotation consistency compared to a single method but this is not a general fact. As expected the use of free keywords and images require tagging that is not directly related to their content, lead to increase the annotators inconsistency.
{"title":"Evaluating Annotators Consistency with the Aid of an Innovative Database Schema","authors":"Zenonas Theodosiou, Olga Georgiou, N. Tsapatsoulis","doi":"10.1109/SMAP.2011.22","DOIUrl":"https://doi.org/10.1109/SMAP.2011.22","url":null,"abstract":"Automatic semantic tagging of multimedia is still inefficient due to the difficulties in modelling abstract or complex terms using low level features. The degree of consensus and homogeneity in judgements among annotators is very important in semantic image and video retrieval. In this paper we present a novel method in evaluating the annotators consistency, which uses an innovative database schema and combines two different annotation approaches. A set of 100 images were annotated by 16 annotators using vocabulary keywords and free keywords. The results indicate that combination of annotation methods may lead to increased annotation consistency compared to a single method but this is not a general fact. As expected the use of free keywords and images require tagging that is not directly related to their content, lead to increase the annotators inconsistency.","PeriodicalId":346975,"journal":{"name":"2011 Sixth International Workshop on Semantic Media Adaptation and Personalization","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2011-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115616010","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}
M. Wallace, Panos Alexopoulos, Ioannis Papafragkos, C. Vassilakis
Manual ontology development is clearly a strenuous task. Whilst a variety of ontological engineering methodologies exist, their actual application is far from trivial, mainly due to the widely diverse nature of the tasks involved. In this work we study these tasks and identify the different types of human experts that are best suited to perform each one. As a result, we present a cooperative version of an special purpose ontological engineering methodology, together with a graphical tool that supports it. Our work is focused on ontology fuzzification, but it can be easily generalized and applied in any ontology engineering context.
{"title":"Supporting User Roles in Ontology Fuzzification","authors":"M. Wallace, Panos Alexopoulos, Ioannis Papafragkos, C. Vassilakis","doi":"10.1109/SMAP.2011.24","DOIUrl":"https://doi.org/10.1109/SMAP.2011.24","url":null,"abstract":"Manual ontology development is clearly a strenuous task. Whilst a variety of ontological engineering methodologies exist, their actual application is far from trivial, mainly due to the widely diverse nature of the tasks involved. In this work we study these tasks and identify the different types of human experts that are best suited to perform each one. As a result, we present a cooperative version of an special purpose ontological engineering methodology, together with a graphical tool that supports it. Our work is focused on ontology fuzzification, but it can be easily generalized and applied in any ontology engineering context.","PeriodicalId":346975,"journal":{"name":"2011 Sixth International Workshop on Semantic Media Adaptation and Personalization","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2011-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130901028","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}
Shallow semantic parsing of natural language processing is an important component in all kind of NLP applications and Semantic Role Labeling in particular, is an active research topic. This paper describes a rule-based Semantic Role Labeling system aimed at extracting semantic information from texts. The input text is processed by exploiting part of speech information and syntactic dependencies in order to identify semantic roles. The system's architecture is presented and the results and further developments are discussed.
{"title":"Extracting Semantic Role Information from Unstructured Texts","authors":"Diana Trandabat, A. Trandabăț","doi":"10.1109/SMAP.2011.20","DOIUrl":"https://doi.org/10.1109/SMAP.2011.20","url":null,"abstract":"Shallow semantic parsing of natural language processing is an important component in all kind of NLP applications and Semantic Role Labeling in particular, is an active research topic. This paper describes a rule-based Semantic Role Labeling system aimed at extracting semantic information from texts. The input text is processed by exploiting part of speech information and syntactic dependencies in order to identify semantic roles. The system's architecture is presented and the results and further developments are discussed.","PeriodicalId":346975,"journal":{"name":"2011 Sixth International Workshop on Semantic Media Adaptation and Personalization","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2011-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132940300","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}
There have been several ontologies developed for context-aware systems and frameworks based on semantic reasoning. In parallel, efforts have been made towards the realisation of ontologies for sensors and observations. While the main inputs of context-aware systems are observations provided by sensors, context ontologies focus only on expressing context data and do not consider the way and in which form this data is retrieved. In this paper, we review dedicated ontologies in the state of the art and bridge this gap by presenting a unifying generic context ontology centred on the concept of entity, together with an integrated sensor and observation ontology linking existing advanced ontologies.
{"title":"Reconciling Context, Observations and Sensors in Ontologies for Pervasive Computing","authors":"Y. Naudet","doi":"10.1109/SMAP.2011.15","DOIUrl":"https://doi.org/10.1109/SMAP.2011.15","url":null,"abstract":"There have been several ontologies developed for context-aware systems and frameworks based on semantic reasoning. In parallel, efforts have been made towards the realisation of ontologies for sensors and observations. While the main inputs of context-aware systems are observations provided by sensors, context ontologies focus only on expressing context data and do not consider the way and in which form this data is retrieved. In this paper, we review dedicated ontologies in the state of the art and bridge this gap by presenting a unifying generic context ontology centred on the concept of entity, together with an integrated sensor and observation ontology linking existing advanced ontologies.","PeriodicalId":346975,"journal":{"name":"2011 Sixth International Workshop on Semantic Media Adaptation and Personalization","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2011-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133753966","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}
Benedita Malheiro, J. Foss, J. C. Burguillo, Ana Peleteiro-Ramallo, Fernando A. Mikic-Fonte
Dynamic personalization of media content is the latest challenge for media content producers and distributors. The idea is to adapt in near real time the content of a video stream to the viewer's profile. This concept encompasses any type of context-awareness customisation, expressed preferences and viewer profiling. To achieve this goal we propose a multi tier framework composed of a content production tier, a content distribution tier and a content consumption tier, representing producers, distributors and viewers, plus an artefact brokerage tier, implemented as an agent-based e brokerage platform, to support the dynamic selection of the content to be inserted in the video stream of each viewer.
{"title":"Dynamic Personalisation of Media Content","authors":"Benedita Malheiro, J. Foss, J. C. Burguillo, Ana Peleteiro-Ramallo, Fernando A. Mikic-Fonte","doi":"10.1109/SMAP.2011.17","DOIUrl":"https://doi.org/10.1109/SMAP.2011.17","url":null,"abstract":"Dynamic personalization of media content is the latest challenge for media content producers and distributors. The idea is to adapt in near real time the content of a video stream to the viewer's profile. This concept encompasses any type of context-awareness customisation, expressed preferences and viewer profiling. To achieve this goal we propose a multi tier framework composed of a content production tier, a content distribution tier and a content consumption tier, representing producers, distributors and viewers, plus an artefact brokerage tier, implemented as an agent-based e brokerage platform, to support the dynamic selection of the content to be inserted in the video stream of each viewer.","PeriodicalId":346975,"journal":{"name":"2011 Sixth International Workshop on Semantic Media Adaptation and Personalization","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2011-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121408342","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}
Maria A. Ioannidou, Eugenia Raptotasiou, I. Anagnostopoulos
We investigate the birth and evolution of the on-line social networks research field as it is outlined in several recent works published within the last four years. Our goal is to present how this particular subject attracts a continually growing interest, as during the last years the on-line social networks are transforming from a simple idea to a social-scientific phenomenon, thus attracting a variety of other related sciences towards the new interesting prospects that emerge on the Web.
{"title":"Social Networking and On-Line Communities: Classification and Research Trends","authors":"Maria A. Ioannidou, Eugenia Raptotasiou, I. Anagnostopoulos","doi":"10.1109/SMAP.2011.19","DOIUrl":"https://doi.org/10.1109/SMAP.2011.19","url":null,"abstract":"We investigate the birth and evolution of the on-line social networks research field as it is outlined in several recent works published within the last four years. Our goal is to present how this particular subject attracts a continually growing interest, as during the last years the on-line social networks are transforming from a simple idea to a social-scientific phenomenon, thus attracting a variety of other related sciences towards the new interesting prospects that emerge on the Web.","PeriodicalId":346975,"journal":{"name":"2011 Sixth International Workshop on Semantic Media Adaptation and Personalization","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2011-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124024218","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}
Data analysis algorithms focused on processing textual data rely on the extraction of relevant features from text and the appropriate association to their formal semantics. In this paper, a method to assist this task, annotating extracted textual features with concepts from a background ontology, is presented. The method is automatic and unsupervised and it has been designed in a generic way, so it can be applied to textual resources ranging from plain text to semi-structured resources (like Wikipedia articles). The system has been tested with tourist destinations and Wikipedia articles showing promising results.
{"title":"A Methodology to Discover Semantic Features from Textual Resources","authors":"C. Vicient, D. Sánchez, Antonio Moreno","doi":"10.1109/SMAP.2011.13","DOIUrl":"https://doi.org/10.1109/SMAP.2011.13","url":null,"abstract":"Data analysis algorithms focused on processing textual data rely on the extraction of relevant features from text and the appropriate association to their formal semantics. In this paper, a method to assist this task, annotating extracted textual features with concepts from a background ontology, is presented. The method is automatic and unsupervised and it has been designed in a generic way, so it can be applied to textual resources ranging from plain text to semi-structured resources (like Wikipedia articles). The system has been tested with tourist destinations and Wikipedia articles showing promising results.","PeriodicalId":346975,"journal":{"name":"2011 Sixth International Workshop on Semantic Media Adaptation and Personalization","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2011-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128802688","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：