Sentiment classification (SC) is an important natural language processing (NLP) task that aims to... more Sentiment classification (SC) is an important natural language processing (NLP) task that aims to determine the sentiment or emotional tone in a given text. With the increasing pervasiveness of internet-based applications and social media, massive amounts of unstructured data are generated daily, elevating the opportunity and challenges associated with automated sentiment extraction for tasks such as customer feedback analysis, social media monitoring, and opinion mining. In this review paper, we provide an update on the state of the art in sentiment analysis, including an overview of and classification methods leveraging machine learning and deep learning methods.
In this article we describe how we apply the concept of coactive emergence as a phenomenon of com... more In this article we describe how we apply the concept of coactive emergence as a phenomenon of complexity that has implications for the design of sensemaking support tools involving a combination of human analysts and software agents. We apply this concept in the design of work methods for distributed sensemaking in cybersecurity work. Sensemaking is a motivated, continuous effort to understand, anticipate, and act upon complex situations. We discuss selected results of a macrocognitive work analysis that informed our focus for design and development of support tools. In that analysis, we identified seven target topics that would be the focus of our research: engaging automation as a full partner, reducing the volume of uncorrelated events, continuous knowledge discovery, more effective visualizations, collaboration and sharing, minimizing tedious work, and architecting scalability and resilience. In addressing the first target topic, we show how coactive emergence inspires an agent-...
In this paper we present a biologically-inspired approach for mission survivability (considered a... more In this paper we present a biologically-inspired approach for mission survivability (considered as the capability of fulfilling a task such as computation) that allows the system to be aware of the possible threats or crises that may arise. This approach uses the notion of resources used by living organisms to control their populations. We present the concept of energetic selection in agent-based evolutionary systems as well as the means to manipulate the configuration of the computation according to the crises or user's specific demands. UTRZYMANIE KRYTYCZNYCH ZADAŃ OPARTE NA ZARZĄDZANIU KRYZYSOWYM W OBLICZENIOWYCH SYSTEMACH WIELOAGENTOWYCH W artykule prezentujemy biologicznie inspirowany mechanizm wspomagający utrzymanie krytycznych zadań (tzw. mission survivability) który umożliwia wykrywanie oraz przeci-wdziałanie wybranym zagrożeniom. Przedstawione podejście wzorowane jest na wykorzysty-waniu przez żywe organizmy zasobów do kontroli populacji. Prezentujemy koncepcje selekcj...
Mapas conceituais sao representacoes graficas do conhecimento de uma pessoa (ou grupo de pessoas)... more Mapas conceituais sao representacoes graficas do conhecimento de uma pessoa (ou grupo de pessoas) sobre um assunto, e como tais, podem ser vistos como um esquema de representacao do conhecimento. Entretanto, a comunidade de Inteligencia Artificial (IA) tende a reprovar o uso do termo “representacao do conhecimento” para mapas conceituais uma vez que estes nao podem ser prontamente traduzidos para uma representacao formal que possa ser utilizada para inferencia e outras tecnicas de IA. Neste trabalho, nos propomos que apesar da flexibilidade de estilo inerentes aos mapas conceituais ha, em mapas bem construidos, caracteristicas especificas (estrutura, semântica, contexto, etc) que permitem o desenvolvimento de ferramentas inteligentes para o auxilio no processo de construcao de mapas conceituais. Argumentamos que a falta de formalismo em favor da flexibilidade proposta pelos mapas conceituais pode ser compensada com a ajuda da IA e de ferramentas inteligentes nos processos de elicita...
Resumo: The most challenging aspect of constructing a concept map is not coming up with the list ... more Resumo: The most challenging aspect of constructing a concept map is not coming up with the list of concepts to include, but linking the concepts into meaningful propositions creating a coherent structure that reflects the learner’s understanding of a domain. We present an algorithm that, during the process of concept mapping, takes the partially constructed map as input to mine the web, and presents to the user a list of suggested concepts that are relevant to the map under construction. Testing a preliminary implementation of the algorithm with a set of users during a concept-mapping workshop seems to validate its viability. Depending on the size of the suggestion list, the algorithm presented on average between 47% and 69% of the concepts in the final maps before the users added them to the map, showing that the algorithm is able to retrieve concepts relevant to the concept mapping effort. Abstract: The most challenging aspect of constructing a concept map is not coming up with t...
Resumo: Atualmente cresce a demanda por treinamento nos diferentes setores da sociedade, onde mai... more Resumo: Atualmente cresce a demanda por treinamento nos diferentes setores da sociedade, onde mais e mais pessoas necessitam adquirir habilidades e conhecimentos de maneira rapida. No entanto, desenvolver um curso contemplando os aspectos tecnicos e pedagogicos nao e uma tarefa facil de ser realizada. Este artigo apresenta uma metodologia para construcao de WBT (Web Based Training) com o intuito de auxiliar os projetistas destes cursos a adicionar qualidade pedagogica no projeto e desenvolvimento destas aplicacoes. Abstract: Atualmente cresce a demanda por treinamento nos diferentes setores da sociedade, onde mais e mais pessoas necessitam adquirir habilidades e conhecimentos de maneira rapida. No entanto, desenvolver um curso contemplando os aspectos tecnicos e pedagogicos nao e uma tarefa facil de ser realizada. Este artigo apresenta uma metodologia para construcao de WBT (Web Based Training) com o intuito de auxiliar os projetistas destes cursos a adicionar qualidade pedagogica n...
Concept maps are a graphical representation of a person's (or group of persons') understa... more Concept maps are a graphical representation of a person's (or group of persons') understanding of a domain. As such, it can be considered a knowledge representation scheme. However, the Artificial Intelligence (AI) community frowns on the use of the term "knowledge representation" to refer to concept maps, because they cannot be readily translated to a formal representation for inference or other AI techniques. In this paper we propose that despite the free-style format that concept maps can take, specific characteristics of well-constructed concept maps (structure, semantics, context, etc.) provide an abundance of information on which to develop smart tools that aid the user in the process of constructing concept maps. Our claim is that the compromise in the formalism in lieu of flexibility proposed by concept maps can be compensated, with the help of AI and smart tools, to help bring the best of both worlds to knowledge elicitation and representation. We demonstr...
With the continuously improving capabilities enabling distributed computing, redundancy and diver... more With the continuously improving capabilities enabling distributed computing, redundancy and diversity of services, Cloud environments are becoming increasingly more attractive for missioncritical and military operations. In such environments, mission assurance and survivability are key enabling factors for deployment, and must be provided as an intrinsic capability of the environment. Mission-critical frameworks must be safe and resistant to localized service failures and compromises. Furthermore, they must be able to autonomously learn and adapt to the environmental challenges and mission requirements. In this paper, we present a biologically inspired approach to mission survivability in cloud computing environments. Our approach introduces a multilayer infrastructure that implements threat detection and service failure coupled with distributed assessments of mission risks, automated re-organization, and re-planning capabilities. Our approach leverages some insights from developmen...
Lecture Notes of the Institute for Computer Sciences, Social Informatics and Telecommunications Engineering, 2017
Spectrum sensing and characterization play a very important role in the implementation of cogniti... more Spectrum sensing and characterization play a very important role in the implementation of cognitive radios and adaptive mobile wireless networks. Most practical mobile network deployments require some level of sensing and adaptation to allow individual nodes to learn and reconfigure based on observations from their own environment. Spectrum sensing can be used for detection of a transmitter in a specific band, which can help cognitive radios to detect spectrum holes for secondary users and to determine the presence of a transmitter in a given area. In addition to determining the existence of a transmitter, information obtained from spectrum sensing can be used to localize a transmitter. In this paper, we focus in oner particular aspect o that problem: the distributed and collaborative sensing, characterization and location of emitters in an open environment. Thus, we propose a software defined radio (SDR)-based spectrum sensing and localization method. The proposed approach uses energy detection for spectrum sensing and fingerprinting techniques for estimating the location of the transmitter. A Universal Software Radio Peripheral (USRP) managed via a small, low-cost computer is used for spectrum sensing. Results obtained from an indoor experimental setup and the K-nearest neighbor algorithm for the fingerprinting based localization are presented in this paper.
MILCOM 2016 - 2016 IEEE Military Communications Conference, 2016
Cyber defense today relies heavily on teams of Subject Matter Experts (SMEs), e.g., Cyber Protect... more Cyber defense today relies heavily on teams of Subject Matter Experts (SMEs), e.g., Cyber Protection Teams (CPTs). Although simple tasks can be automated or scripted, complex decision processes-increasingly needed to counter cyber threats-require SME insight and manual execution. As a result, cyber-defense operations tend to emphasize collection and archiving of data over real-time decision making and response, postponing actionable analysis and response until later, where “later” is frequently “too late.” In contrast, adversaries are readily using automation tools to minimize manual work and encapsulate autonomous behaviors into botnets and viruses that adapt to changing conditions. This imbalance puts the adversary in a position of advantage, a situation the research presented in this paper aims to remedy. The scarcity of cyber SMEs and the high cost of involving them in manual cyber responses are among the main factors contributing to the imbalance. The approach we describe aims to reduce the reliance on human SMEs, drive down the cost, and increase the effectiveness of CPTs by capturing expert knowledge in a tool that will automate the identification of known and unknown threats and the launching of mitigations to counter ongoing attacks at system speeds.
Sentiment classification (SC) is an important natural language processing (NLP) task that aims to... more Sentiment classification (SC) is an important natural language processing (NLP) task that aims to determine the sentiment or emotional tone in a given text. With the increasing pervasiveness of internet-based applications and social media, massive amounts of unstructured data are generated daily, elevating the opportunity and challenges associated with automated sentiment extraction for tasks such as customer feedback analysis, social media monitoring, and opinion mining. In this review paper, we provide an update on the state of the art in sentiment analysis, including an overview of and classification methods leveraging machine learning and deep learning methods.
In this article we describe how we apply the concept of coactive emergence as a phenomenon of com... more In this article we describe how we apply the concept of coactive emergence as a phenomenon of complexity that has implications for the design of sensemaking support tools involving a combination of human analysts and software agents. We apply this concept in the design of work methods for distributed sensemaking in cybersecurity work. Sensemaking is a motivated, continuous effort to understand, anticipate, and act upon complex situations. We discuss selected results of a macrocognitive work analysis that informed our focus for design and development of support tools. In that analysis, we identified seven target topics that would be the focus of our research: engaging automation as a full partner, reducing the volume of uncorrelated events, continuous knowledge discovery, more effective visualizations, collaboration and sharing, minimizing tedious work, and architecting scalability and resilience. In addressing the first target topic, we show how coactive emergence inspires an agent-...
In this paper we present a biologically-inspired approach for mission survivability (considered a... more In this paper we present a biologically-inspired approach for mission survivability (considered as the capability of fulfilling a task such as computation) that allows the system to be aware of the possible threats or crises that may arise. This approach uses the notion of resources used by living organisms to control their populations. We present the concept of energetic selection in agent-based evolutionary systems as well as the means to manipulate the configuration of the computation according to the crises or user's specific demands. UTRZYMANIE KRYTYCZNYCH ZADAŃ OPARTE NA ZARZĄDZANIU KRYZYSOWYM W OBLICZENIOWYCH SYSTEMACH WIELOAGENTOWYCH W artykule prezentujemy biologicznie inspirowany mechanizm wspomagający utrzymanie krytycznych zadań (tzw. mission survivability) który umożliwia wykrywanie oraz przeci-wdziałanie wybranym zagrożeniom. Przedstawione podejście wzorowane jest na wykorzysty-waniu przez żywe organizmy zasobów do kontroli populacji. Prezentujemy koncepcje selekcj...
Mapas conceituais sao representacoes graficas do conhecimento de uma pessoa (ou grupo de pessoas)... more Mapas conceituais sao representacoes graficas do conhecimento de uma pessoa (ou grupo de pessoas) sobre um assunto, e como tais, podem ser vistos como um esquema de representacao do conhecimento. Entretanto, a comunidade de Inteligencia Artificial (IA) tende a reprovar o uso do termo “representacao do conhecimento” para mapas conceituais uma vez que estes nao podem ser prontamente traduzidos para uma representacao formal que possa ser utilizada para inferencia e outras tecnicas de IA. Neste trabalho, nos propomos que apesar da flexibilidade de estilo inerentes aos mapas conceituais ha, em mapas bem construidos, caracteristicas especificas (estrutura, semântica, contexto, etc) que permitem o desenvolvimento de ferramentas inteligentes para o auxilio no processo de construcao de mapas conceituais. Argumentamos que a falta de formalismo em favor da flexibilidade proposta pelos mapas conceituais pode ser compensada com a ajuda da IA e de ferramentas inteligentes nos processos de elicita...
Resumo: The most challenging aspect of constructing a concept map is not coming up with the list ... more Resumo: The most challenging aspect of constructing a concept map is not coming up with the list of concepts to include, but linking the concepts into meaningful propositions creating a coherent structure that reflects the learner’s understanding of a domain. We present an algorithm that, during the process of concept mapping, takes the partially constructed map as input to mine the web, and presents to the user a list of suggested concepts that are relevant to the map under construction. Testing a preliminary implementation of the algorithm with a set of users during a concept-mapping workshop seems to validate its viability. Depending on the size of the suggestion list, the algorithm presented on average between 47% and 69% of the concepts in the final maps before the users added them to the map, showing that the algorithm is able to retrieve concepts relevant to the concept mapping effort. Abstract: The most challenging aspect of constructing a concept map is not coming up with t...
Resumo: Atualmente cresce a demanda por treinamento nos diferentes setores da sociedade, onde mai... more Resumo: Atualmente cresce a demanda por treinamento nos diferentes setores da sociedade, onde mais e mais pessoas necessitam adquirir habilidades e conhecimentos de maneira rapida. No entanto, desenvolver um curso contemplando os aspectos tecnicos e pedagogicos nao e uma tarefa facil de ser realizada. Este artigo apresenta uma metodologia para construcao de WBT (Web Based Training) com o intuito de auxiliar os projetistas destes cursos a adicionar qualidade pedagogica no projeto e desenvolvimento destas aplicacoes. Abstract: Atualmente cresce a demanda por treinamento nos diferentes setores da sociedade, onde mais e mais pessoas necessitam adquirir habilidades e conhecimentos de maneira rapida. No entanto, desenvolver um curso contemplando os aspectos tecnicos e pedagogicos nao e uma tarefa facil de ser realizada. Este artigo apresenta uma metodologia para construcao de WBT (Web Based Training) com o intuito de auxiliar os projetistas destes cursos a adicionar qualidade pedagogica n...
Concept maps are a graphical representation of a person's (or group of persons') understa... more Concept maps are a graphical representation of a person's (or group of persons') understanding of a domain. As such, it can be considered a knowledge representation scheme. However, the Artificial Intelligence (AI) community frowns on the use of the term "knowledge representation" to refer to concept maps, because they cannot be readily translated to a formal representation for inference or other AI techniques. In this paper we propose that despite the free-style format that concept maps can take, specific characteristics of well-constructed concept maps (structure, semantics, context, etc.) provide an abundance of information on which to develop smart tools that aid the user in the process of constructing concept maps. Our claim is that the compromise in the formalism in lieu of flexibility proposed by concept maps can be compensated, with the help of AI and smart tools, to help bring the best of both worlds to knowledge elicitation and representation. We demonstr...
With the continuously improving capabilities enabling distributed computing, redundancy and diver... more With the continuously improving capabilities enabling distributed computing, redundancy and diversity of services, Cloud environments are becoming increasingly more attractive for missioncritical and military operations. In such environments, mission assurance and survivability are key enabling factors for deployment, and must be provided as an intrinsic capability of the environment. Mission-critical frameworks must be safe and resistant to localized service failures and compromises. Furthermore, they must be able to autonomously learn and adapt to the environmental challenges and mission requirements. In this paper, we present a biologically inspired approach to mission survivability in cloud computing environments. Our approach introduces a multilayer infrastructure that implements threat detection and service failure coupled with distributed assessments of mission risks, automated re-organization, and re-planning capabilities. Our approach leverages some insights from developmen...
Lecture Notes of the Institute for Computer Sciences, Social Informatics and Telecommunications Engineering, 2017
Spectrum sensing and characterization play a very important role in the implementation of cogniti... more Spectrum sensing and characterization play a very important role in the implementation of cognitive radios and adaptive mobile wireless networks. Most practical mobile network deployments require some level of sensing and adaptation to allow individual nodes to learn and reconfigure based on observations from their own environment. Spectrum sensing can be used for detection of a transmitter in a specific band, which can help cognitive radios to detect spectrum holes for secondary users and to determine the presence of a transmitter in a given area. In addition to determining the existence of a transmitter, information obtained from spectrum sensing can be used to localize a transmitter. In this paper, we focus in oner particular aspect o that problem: the distributed and collaborative sensing, characterization and location of emitters in an open environment. Thus, we propose a software defined radio (SDR)-based spectrum sensing and localization method. The proposed approach uses energy detection for spectrum sensing and fingerprinting techniques for estimating the location of the transmitter. A Universal Software Radio Peripheral (USRP) managed via a small, low-cost computer is used for spectrum sensing. Results obtained from an indoor experimental setup and the K-nearest neighbor algorithm for the fingerprinting based localization are presented in this paper.
MILCOM 2016 - 2016 IEEE Military Communications Conference, 2016
Cyber defense today relies heavily on teams of Subject Matter Experts (SMEs), e.g., Cyber Protect... more Cyber defense today relies heavily on teams of Subject Matter Experts (SMEs), e.g., Cyber Protection Teams (CPTs). Although simple tasks can be automated or scripted, complex decision processes-increasingly needed to counter cyber threats-require SME insight and manual execution. As a result, cyber-defense operations tend to emphasize collection and archiving of data over real-time decision making and response, postponing actionable analysis and response until later, where “later” is frequently “too late.” In contrast, adversaries are readily using automation tools to minimize manual work and encapsulate autonomous behaviors into botnets and viruses that adapt to changing conditions. This imbalance puts the adversary in a position of advantage, a situation the research presented in this paper aims to remedy. The scarcity of cyber SMEs and the high cost of involving them in manual cyber responses are among the main factors contributing to the imbalance. The approach we describe aims to reduce the reliance on human SMEs, drive down the cost, and increase the effectiveness of CPTs by capturing expert knowledge in a tool that will automate the identification of known and unknown threats and the launching of mitigations to counter ongoing attacks at system speeds.
Autonomous intelligent agents are often used to command autonomous systems in their mission execu... more Autonomous intelligent agents are often used to command autonomous systems in their mission execution. These intelligent agents can be implemented in a cognitive architecture in order to perform human-like reasoning about the mission. In this design approach the agent can be modeled by associating behaviors with the components present in the cognitive architecture. One of the architectures we investigate in our research effort is the rule-based reasoning system, Soar. In this framework, agents input data to form the perception of the situation, and then use Soar rules to make decisions and propose new actions. But the rule-based representations in cognitive architectures like Soar are not amenable to formal, rigorous analysis. Understanding when the cognitive architecture will deviate from the desired behavior due to environment or adaptation requires creating and proving adherence to a formal specification of acceptability. There is a critical need for verification methods to be integrated with the design and operation of intelligent agents to assure correct execution before deploying in safety critical missions. In this work we evaluate verification by model transformation focused on one class of cognitive architectures translated into an analytical domain. This translation enables analysis of the complete set of rules in the cognitive architecture to identify if there exists any conflict among the rules, violation of safety properties, or reduction in performance characteristics. The rules used in the Soar cognitive architecture were transformed into the format of the Uppaal real-time verification tool. Using Uppaal, we were able to verify temporal logic properties about the behavior of the autonomous agent. This involved properties such as completion of action sequences and appropriate action selection. This was demonstrated in a prototype system that performed in-flight checklist maintenance. The translation from a rule-based representation to a verifiable representation accomplishes an important first step in achieving verifiable operation of adaptive decision-making computational agents. As adaptive computational agents gain more ground in cooperative and autonomous driving cars, unmanned aerial vehicles, and production equipment, the ability to constrain the operation of these agents within a formally specified and well-understood set of performance and
Uploads
Papers by Marco Carvalho