Pub Date : 2016-11-01DOI: 10.1109/AICCSA.2016.7945816
Jisun An, Haewoon Kwak, B. Jansen
Using personas during interactive design has considerable potential for product and content development. Unfortunately, personas have typically been a fairly static technique. In this research, we validate an approach for creating personas in real time, based on analysis of actual social media data in an effort to automate the generation of personas. We validate that social media data can be implemented as an approach for automating generating personas in real time using actual YouTube social media data from a global media corporation that produces online digital content. Using the organization's YouTube channel, we collect demographic data, customer interactions, and topical interests, leveraging more than 188,000 subscriber profiles and more than 30 million user interactions. Then, we conduct statistical analysis on the social media data to determine whether the data could lead to the generation of valid personas based on statistically difference market segments. Findings show that customers can be segmented using product topics by gender and age based using social media data. However, our findings also show that the data is biased by the content created. The results offer insights into competitive marketing and product preferences for the consumers of the online digital content. Implications are that personas can be generated in real-time using social media data, instead of a time-consuming manual development process.
{"title":"Validating social media data for automatic persona generation","authors":"Jisun An, Haewoon Kwak, B. Jansen","doi":"10.1109/AICCSA.2016.7945816","DOIUrl":"https://doi.org/10.1109/AICCSA.2016.7945816","url":null,"abstract":"Using personas during interactive design has considerable potential for product and content development. Unfortunately, personas have typically been a fairly static technique. In this research, we validate an approach for creating personas in real time, based on analysis of actual social media data in an effort to automate the generation of personas. We validate that social media data can be implemented as an approach for automating generating personas in real time using actual YouTube social media data from a global media corporation that produces online digital content. Using the organization's YouTube channel, we collect demographic data, customer interactions, and topical interests, leveraging more than 188,000 subscriber profiles and more than 30 million user interactions. Then, we conduct statistical analysis on the social media data to determine whether the data could lead to the generation of valid personas based on statistically difference market segments. Findings show that customers can be segmented using product topics by gender and age based using social media data. However, our findings also show that the data is biased by the content created. The results offer insights into competitive marketing and product preferences for the consumers of the online digital content. Implications are that personas can be generated in real-time using social media data, instead of a time-consuming manual development process.","PeriodicalId":448329,"journal":{"name":"2016 IEEE/ACS 13th International Conference of Computer Systems and Applications (AICCSA)","volume":"14 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126804840","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}
Pub Date : 2016-11-01DOI: 10.1109/AICCSA.2016.7945762
S. Handouf, Sara Arabi, Essaid Sabir, M. Sadik
This paper presents some interesting findings about consumer confusion vis-a-vis real price for offered services and its impact on Telecommunications market dynamics. Noting that pricing strategies have a critical influence on consumer choice, customer satisfaction and customer retention in the telecommunications sector, our study focuses on the analysis of the service providers' strategic behavior. Yet, we construct a simple oligopolistic model to capture the interactions among service providers and end users. Next, we analyze the behavior of service providers in terms of their pricing strategies within the framework of non-cooperative game theory. It aims to evaluate the impact of consumer confusion on the competition and operators profitability. Here, rational service providers are competing with each other to maximize their respective payoffs in presence of both a fraction of confused consumers while others are non confused. We provide some interesting results regarding the Nash equilibrium of this game. More precisely, we showed existence and uniqueness of the Nash equilibrium under some conditions. Furthermore, we introduce two learning algorithms that may lead operators to learn their strategic pricing schemes in a complete distributed manner. Extensive simulations show convergence of a proposed scheme to the Nash equilibrium and give some insights on how the game parameters may vary the oligopoly outcome.
{"title":"Telecommunication market share game: Inducing boundedly rational consumers via price misperception","authors":"S. Handouf, Sara Arabi, Essaid Sabir, M. Sadik","doi":"10.1109/AICCSA.2016.7945762","DOIUrl":"https://doi.org/10.1109/AICCSA.2016.7945762","url":null,"abstract":"This paper presents some interesting findings about consumer confusion vis-a-vis real price for offered services and its impact on Telecommunications market dynamics. Noting that pricing strategies have a critical influence on consumer choice, customer satisfaction and customer retention in the telecommunications sector, our study focuses on the analysis of the service providers' strategic behavior. Yet, we construct a simple oligopolistic model to capture the interactions among service providers and end users. Next, we analyze the behavior of service providers in terms of their pricing strategies within the framework of non-cooperative game theory. It aims to evaluate the impact of consumer confusion on the competition and operators profitability. Here, rational service providers are competing with each other to maximize their respective payoffs in presence of both a fraction of confused consumers while others are non confused. We provide some interesting results regarding the Nash equilibrium of this game. More precisely, we showed existence and uniqueness of the Nash equilibrium under some conditions. Furthermore, we introduce two learning algorithms that may lead operators to learn their strategic pricing schemes in a complete distributed manner. Extensive simulations show convergence of a proposed scheme to the Nash equilibrium and give some insights on how the game parameters may vary the oligopoly outcome.","PeriodicalId":448329,"journal":{"name":"2016 IEEE/ACS 13th International Conference of Computer Systems and Applications (AICCSA)","volume":"22 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124322762","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}
Pub Date : 2016-11-01DOI: 10.1109/AICCSA.2016.7945676
Matthieu Mastio, Mahdi Zargayouna, G. Scémama, O. Rana
Travelers mobility simulation is a powerful tool to test strategies in a virtual environment, without impacting the quality of the real traffic network. However, existing mobility multiagent and micro-simulations can only consider a sample of the real volumes of travelers, especially for big regions. With distributed simulations, it would be easier to analyze and predict the status of nowadays networks. This kind of simulations requires big computational power and methods to split the simulation between several machines. This work describes how to achieve such a distribution in a microscopic simulation context, and compare our results with a previous work on macroscopic simulation.
{"title":"Patterns to distribute mobility simulations","authors":"Matthieu Mastio, Mahdi Zargayouna, G. Scémama, O. Rana","doi":"10.1109/AICCSA.2016.7945676","DOIUrl":"https://doi.org/10.1109/AICCSA.2016.7945676","url":null,"abstract":"Travelers mobility simulation is a powerful tool to test strategies in a virtual environment, without impacting the quality of the real traffic network. However, existing mobility multiagent and micro-simulations can only consider a sample of the real volumes of travelers, especially for big regions. With distributed simulations, it would be easier to analyze and predict the status of nowadays networks. This kind of simulations requires big computational power and methods to split the simulation between several machines. This work describes how to achieve such a distribution in a microscopic simulation context, and compare our results with a previous work on macroscopic simulation.","PeriodicalId":448329,"journal":{"name":"2016 IEEE/ACS 13th International Conference of Computer Systems and Applications (AICCSA)","volume":"103 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123624954","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}
Pub Date : 2016-11-01DOI: 10.1109/AICCSA.2016.7945794
Emna Ben Ayed, C. Kolski, H. Ezzedine
The technological revolution has been accompanied with the appearance of mobile systems. Mobile Knowledge Management Systems (KMS) which introduce a new style of managing organizations using mobile devices, are used in different contexts of use. So, a context based evaluation of such systems becomes required. Our contribution consists of enhancing the evaluation of mobile KMS that was previously confined to the classifier performance evaluation. In this work, we propose a quality model which ensures a global evaluation of mobile KMS. The proposed quality model evaluates the mobile KMS regarding three dimensions: classifier performance, usability and quality of promotion. It defines a set of criteria and allows their measurement. To validate the proposed approach, an evaluation system has been developed. The results obtained from a series of evaluation results drawing on 20 users show that the continuous change of the context affects the use of a mobile KMS.
{"title":"Context-based evaluation of mobile knowledge management systems","authors":"Emna Ben Ayed, C. Kolski, H. Ezzedine","doi":"10.1109/AICCSA.2016.7945794","DOIUrl":"https://doi.org/10.1109/AICCSA.2016.7945794","url":null,"abstract":"The technological revolution has been accompanied with the appearance of mobile systems. Mobile Knowledge Management Systems (KMS) which introduce a new style of managing organizations using mobile devices, are used in different contexts of use. So, a context based evaluation of such systems becomes required. Our contribution consists of enhancing the evaluation of mobile KMS that was previously confined to the classifier performance evaluation. In this work, we propose a quality model which ensures a global evaluation of mobile KMS. The proposed quality model evaluates the mobile KMS regarding three dimensions: classifier performance, usability and quality of promotion. It defines a set of criteria and allows their measurement. To validate the proposed approach, an evaluation system has been developed. The results obtained from a series of evaluation results drawing on 20 users show that the continuous change of the context affects the use of a mobile KMS.","PeriodicalId":448329,"journal":{"name":"2016 IEEE/ACS 13th International Conference of Computer Systems and Applications (AICCSA)","volume":"27 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125501061","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}
Pub Date : 2016-11-01DOI: 10.1109/AICCSA.2016.7945683
Ahmed Aboud, Haifa Touati
Named Data Networking (NDN) has emerged as a new and promising information-centric architecture for the future Internet that is also gaining momentum in Wireless Sensors Networks (WSNs) as an alternative paradigm to traditional IP. However, the main solutions in current works lack energy efficient design to meet the severely limited energy resources in WSNs. Moreover, the NDN architecture cannot natively allow devices to transmit unsolicited data, like alarms or status changes unless to properly modify the semantics of exchanged packets and the forwarding strategy. In this paper, we propose a Geographic Interest Forwarding scheme called GIF in short, for NDN-Based WSNs. In this proposal, we added support for push-based WSN traffic. We further extend the scheme with an efficient forwarding techniques in which several energy efficient mechanisms including flooding scope control, broadcast storm avoidance, packet suppression are designed to balance the energy consumption across the network. We extend the ndnSIM to support wireless multihop communication to validate the proposed scheme. Simulation experiments confirm the viability and effectiveness of the proposal.
{"title":"Geographic interest forwarding in NDN-based wireless sensor networks","authors":"Ahmed Aboud, Haifa Touati","doi":"10.1109/AICCSA.2016.7945683","DOIUrl":"https://doi.org/10.1109/AICCSA.2016.7945683","url":null,"abstract":"Named Data Networking (NDN) has emerged as a new and promising information-centric architecture for the future Internet that is also gaining momentum in Wireless Sensors Networks (WSNs) as an alternative paradigm to traditional IP. However, the main solutions in current works lack energy efficient design to meet the severely limited energy resources in WSNs. Moreover, the NDN architecture cannot natively allow devices to transmit unsolicited data, like alarms or status changes unless to properly modify the semantics of exchanged packets and the forwarding strategy. In this paper, we propose a Geographic Interest Forwarding scheme called GIF in short, for NDN-Based WSNs. In this proposal, we added support for push-based WSN traffic. We further extend the scheme with an efficient forwarding techniques in which several energy efficient mechanisms including flooding scope control, broadcast storm avoidance, packet suppression are designed to balance the energy consumption across the network. We extend the ndnSIM to support wireless multihop communication to validate the proposed scheme. Simulation experiments confirm the viability and effectiveness of the proposal.","PeriodicalId":448329,"journal":{"name":"2016 IEEE/ACS 13th International Conference of Computer Systems and Applications (AICCSA)","volume":"44 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125627940","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}
Pub Date : 2016-11-01DOI: 10.1109/AICCSA.2016.7945700
F. Z. Ennaji, A. E. Fazziki, Hasna El Alaouiel El Abdallaoui, Abderrahmane Sadiq, M. Sadgal, D. Benslimane
Numerous studies have discussed the benefits of using social networks, even companies started to exploit the usefulness of this valuable information sources. Collecting social data then integrating them into a CRM (Customer Relationship Management) has led companies to understand the customer needs and therefore to improve the development process of their products or their services quality. In this work, we propose a multi-agent framework for analyzing extracted opinions from social media. In the development process, we were brought to consider the huge volumes of data (Big Data) and the response time. To do so, an architecture based on Map/Reduce analysis using Hadoop was made in order to perform the data refinement (classify or remove special words or delete the unvaluable reviews) and sentiment analysis (Sentigem). Finally, a study case using Twitter (Twitter4J API) as a data source, was made to verify the effectiveness of the proposed framework.
{"title":"Multi-agent framework for social CRM: Extracting and analyzing opinions","authors":"F. Z. Ennaji, A. E. Fazziki, Hasna El Alaouiel El Abdallaoui, Abderrahmane Sadiq, M. Sadgal, D. Benslimane","doi":"10.1109/AICCSA.2016.7945700","DOIUrl":"https://doi.org/10.1109/AICCSA.2016.7945700","url":null,"abstract":"Numerous studies have discussed the benefits of using social networks, even companies started to exploit the usefulness of this valuable information sources. Collecting social data then integrating them into a CRM (Customer Relationship Management) has led companies to understand the customer needs and therefore to improve the development process of their products or their services quality. In this work, we propose a multi-agent framework for analyzing extracted opinions from social media. In the development process, we were brought to consider the huge volumes of data (Big Data) and the response time. To do so, an architecture based on Map/Reduce analysis using Hadoop was made in order to perform the data refinement (classify or remove special words or delete the unvaluable reviews) and sentiment analysis (Sentigem). Finally, a study case using Twitter (Twitter4J API) as a data source, was made to verify the effectiveness of the proposed framework.","PeriodicalId":448329,"journal":{"name":"2016 IEEE/ACS 13th International Conference of Computer Systems and Applications (AICCSA)","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131695868","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}
Pub Date : 2016-11-01DOI: 10.1109/AICCSA.2016.7945714
A. Alaboudi, A. Atkins, B. Sharp, M. Alzahrani, Ahmed Balkhair, Tamara Sunbul
Regardless of promised benefits of telemedicine and its potential technical superiority, its adoption is often cited as a failed project and the lack of understanding potential user acceptance is one of the most pervasive reasons leading to this failure. More than 40% of Information and Communications Technology (ICT) adoption projects have failed for this reason and any ICT system would not be effective if its potential user were averse to utilising its adoption. To the best of our knowledge, no comprehensive scientific study has investigated the perceptions and attitudes of clinical staff towards telemedicine acceptance and use in the Kingdom of Saudi Arabia (KSA) healthcare system at a national level. The objective of this novel research lies in addressing this gap and is conducted in collaboration with the Saudi Ministry of Health (MOH). The Unified Theory of Acceptance and Use of Technology (UTAUT) is used to underpin this research and a questionnaire to support the objective of this research was conducted in KSA to collect data from a random sample of different categories of clinical staff in KSA. The survey of 1,523 returned questionnaires were selected and the findings of this study show that 67.6% of clinical staff on average expect or extremely expect high performance from using telemedicine in KSA. Also 62.6% of clinical staff agree or extremely agree that the use of telemedicine in KSA will be compatible with culture and society. On the other hand, 38.0% of clinical staff agree or extremely agree that the use of telemedicine will not be compatible with their job role. Furthermore 40.6% of participants stated that using telemedicine would not be easy for clinical staff and 37.4% of them indicated that it will waste a lot of time and require a lot of technical and non-clinical administrative time to complete
{"title":"Perceptions and attitudes of clinical staff towards telemedicine acceptance in Saudi Arabia","authors":"A. Alaboudi, A. Atkins, B. Sharp, M. Alzahrani, Ahmed Balkhair, Tamara Sunbul","doi":"10.1109/AICCSA.2016.7945714","DOIUrl":"https://doi.org/10.1109/AICCSA.2016.7945714","url":null,"abstract":"Regardless of promised benefits of telemedicine and its potential technical superiority, its adoption is often cited as a failed project and the lack of understanding potential user acceptance is one of the most pervasive reasons leading to this failure. More than 40% of Information and Communications Technology (ICT) adoption projects have failed for this reason and any ICT system would not be effective if its potential user were averse to utilising its adoption. To the best of our knowledge, no comprehensive scientific study has investigated the perceptions and attitudes of clinical staff towards telemedicine acceptance and use in the Kingdom of Saudi Arabia (KSA) healthcare system at a national level. The objective of this novel research lies in addressing this gap and is conducted in collaboration with the Saudi Ministry of Health (MOH). The Unified Theory of Acceptance and Use of Technology (UTAUT) is used to underpin this research and a questionnaire to support the objective of this research was conducted in KSA to collect data from a random sample of different categories of clinical staff in KSA. The survey of 1,523 returned questionnaires were selected and the findings of this study show that 67.6% of clinical staff on average expect or extremely expect high performance from using telemedicine in KSA. Also 62.6% of clinical staff agree or extremely agree that the use of telemedicine in KSA will be compatible with culture and society. On the other hand, 38.0% of clinical staff agree or extremely agree that the use of telemedicine will not be compatible with their job role. Furthermore 40.6% of participants stated that using telemedicine would not be easy for clinical staff and 37.4% of them indicated that it will waste a lot of time and require a lot of technical and non-clinical administrative time to complete","PeriodicalId":448329,"journal":{"name":"2016 IEEE/ACS 13th International Conference of Computer Systems and Applications (AICCSA)","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122193579","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}
Pub Date : 2016-11-01DOI: 10.1109/AICCSA.2016.7945829
Ichrak Mehrez, O. Hamdi-Larbi, T. Dufaud, N. Emad
Several applications in numerical scientific computing process sparse matrices with either a regular or irregular structure. The very large size of these matrices requires to use compressing formats and target parallel/distributed architectures in order to reduce both space complexity and processing time. The optimal compression format (OCF) of such matrices may in fact vary according to both the application context of the numerical method and the target hardware architecture. In this paper, we propose a design of a system that automatically selects the OCF according to the two above cited parameters. The expert system obtained from our model targets dynamic integration of the user expertise thus allowing better performances. The optimal format selection is based on the makespan criterion. As a first validation test of our system, we studied the representative case of Horner scheme in the context of data parallel programming model and multicore cluster. Our experiments focus on the four compression formats CSR, CSC, COO and ELLPACK and their complexities in a data parallel programming model context.
{"title":"Towards an auto-tuning system design for optimal sparse compression format selection with user expertise","authors":"Ichrak Mehrez, O. Hamdi-Larbi, T. Dufaud, N. Emad","doi":"10.1109/AICCSA.2016.7945829","DOIUrl":"https://doi.org/10.1109/AICCSA.2016.7945829","url":null,"abstract":"Several applications in numerical scientific computing process sparse matrices with either a regular or irregular structure. The very large size of these matrices requires to use compressing formats and target parallel/distributed architectures in order to reduce both space complexity and processing time. The optimal compression format (OCF) of such matrices may in fact vary according to both the application context of the numerical method and the target hardware architecture. In this paper, we propose a design of a system that automatically selects the OCF according to the two above cited parameters. The expert system obtained from our model targets dynamic integration of the user expertise thus allowing better performances. The optimal format selection is based on the makespan criterion. As a first validation test of our system, we studied the representative case of Horner scheme in the context of data parallel programming model and multicore cluster. Our experiments focus on the four compression formats CSR, CSC, COO and ELLPACK and their complexities in a data parallel programming model context.","PeriodicalId":448329,"journal":{"name":"2016 IEEE/ACS 13th International Conference of Computer Systems and Applications (AICCSA)","volume":"102 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129818182","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}
Pub Date : 2016-11-01DOI: 10.1109/AICCSA.2016.7945629
Ayman El-Kilany, N. Tazi, Ehab Ezzat
A considerable amount of attributes in real datasets are not numerical, but rather textual and categorical. We investigate the problem of identifying outliers in categorical and textual datasets. We propose a clustering-based semi-supervised outlier detection method which basically represents normal and unlabeled data points as a bipartite graph. We leverage the existing free of parameters clustering techniques to cluster the resulting graph. The bipartite graph is clustered with a specific end goal to distinguish unlabeled data points as either outliers or normal data points. The proposed method is evaluated using multiple categorical and textual datasets against one-class support vector machines classifier and FRaC approach for semi-supervised outlier detection where the proposed method has shown a comparable performance.
{"title":"Semi-supervised outlier detection via bipartite graph clustering","authors":"Ayman El-Kilany, N. Tazi, Ehab Ezzat","doi":"10.1109/AICCSA.2016.7945629","DOIUrl":"https://doi.org/10.1109/AICCSA.2016.7945629","url":null,"abstract":"A considerable amount of attributes in real datasets are not numerical, but rather textual and categorical. We investigate the problem of identifying outliers in categorical and textual datasets. We propose a clustering-based semi-supervised outlier detection method which basically represents normal and unlabeled data points as a bipartite graph. We leverage the existing free of parameters clustering techniques to cluster the resulting graph. The bipartite graph is clustered with a specific end goal to distinguish unlabeled data points as either outliers or normal data points. The proposed method is evaluated using multiple categorical and textual datasets against one-class support vector machines classifier and FRaC approach for semi-supervised outlier detection where the proposed method has shown a comparable performance.","PeriodicalId":448329,"journal":{"name":"2016 IEEE/ACS 13th International Conference of Computer Systems and Applications (AICCSA)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129993570","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}
Pub Date : 2016-11-01DOI: 10.1109/AICCSA.2016.7945644
Houssem Turki, Mohamed Ben Halima, A. Alimi
Text detection in natural scenes holds great importance in the field of research and still remains a challenge because of size, various fonts, line orientation, different illumination conditions, weak character and complex background in image. The contribution of the proposed method is filtering out complex backgrounds by utilizing two masks filtering based on text confidence map in the first step and multi-channel maximally stable extremal regions (MSERs) in the second step. Both steps are designed to enhancement, maximize capacity of zones text pixels candidates to distinguish text boxes from the rest of the image. Then non-text components are filtered by the classification of character candidate based on Support Vector Machines (SVM) using HOG features. The false positives are eliminated by geometrical properties of text blocks. Finally we apply boundary box localization after a stage of word grouping. The proposed method has been evaluated on ICDAR 2013 scene text detection competition dataset and the encouraging experiments results demonstrate the robustness of our method.
{"title":"Text detection in natural scene images using two masks filtering","authors":"Houssem Turki, Mohamed Ben Halima, A. Alimi","doi":"10.1109/AICCSA.2016.7945644","DOIUrl":"https://doi.org/10.1109/AICCSA.2016.7945644","url":null,"abstract":"Text detection in natural scenes holds great importance in the field of research and still remains a challenge because of size, various fonts, line orientation, different illumination conditions, weak character and complex background in image. The contribution of the proposed method is filtering out complex backgrounds by utilizing two masks filtering based on text confidence map in the first step and multi-channel maximally stable extremal regions (MSERs) in the second step. Both steps are designed to enhancement, maximize capacity of zones text pixels candidates to distinguish text boxes from the rest of the image. Then non-text components are filtered by the classification of character candidate based on Support Vector Machines (SVM) using HOG features. The false positives are eliminated by geometrical properties of text blocks. Finally we apply boundary box localization after a stage of word grouping. The proposed method has been evaluated on ICDAR 2013 scene text detection competition dataset and the encouraging experiments results demonstrate the robustness of our method.","PeriodicalId":448329,"journal":{"name":"2016 IEEE/ACS 13th International Conference of Computer Systems and Applications (AICCSA)","volume":"39 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130273643","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：