Adina COCU | University Dunarea de Jos of Galati (original) (raw)
Papers by Adina COCU
ICERI proceedings, Nov 1, 2022
Lecture notes in networks and systems, 2023
THE ANNALS OF “DUNAREA DE JOS” UNIVERSITY OF GALATI FASCICLE III, 2017, VOL. 40, NO. 1, ISSN 2344... more THE ANNALS OF “DUNAREA DE JOS” UNIVERSITY OF GALATI FASCICLE III, 2017, VOL. 40, NO. 1, ISSN 2344-4738, ISSN-L 1221-454X ELECTROTECHNICS, ELECTRONICS, AUTOMATIC CONTROL, INFORMATICS
The European Proceedings of Social and Behavioural Sciences, 2019
This is an Open Access article distributed under the terms of the Creative Commons Attribution-No... more This is an Open Access article distributed under the terms of the Creative Commons Attribution-Noncommercial 4.0 Unported License, permitting all non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
There are several approaches in trying to solve the Quantitative 1Structure-Activity Relationship... more There are several approaches in trying to solve the Quantitative 1Structure-Activity Relationship (QSAR) problem. These approaches are based either on statistical methods or on predictive data mining. Among the statistical methods, one should consider regression analysis, pattern recognition (such as cluster analysis, factor analysis and principal components analysis) or partial least squares. Predictive data mining techniques use either neural networks, or genetic programming, or neuro-fuzzy knowledge. These approaches have a low explanatory capability or non at all. This paper attempts to establish a new approach in solving QSAR problems using descriptive data mining. This way, the relationship between the chemical properties and the activity of a substance would be comprehensibly modeled.
Abstract: In intelligent tutoring systems, student or user modeling implies dealing with imperfec... more Abstract: In intelligent tutoring systems, student or user modeling implies dealing with imperfect and uncertain knowledge. One of the artificial intelligence techniques used for uncertainty management is that of Bayesian networks. This paradigm is recommended in the situation when exist dependencies between data and qualitative information about these data. In this work we present a student knowledge diagnosis model based on representation with Bayesian networks. The educational system incorporate a multimedia interface for accomplishes the testing tools. The results of testing sessions are represented and interpreted with probability theory in order to ensure an adapted support for the student. The aims of the computer assisted application that contains this diagnose module are to support the student in personalized learning process and errors explanation.
Education is ‐ clearly ‐ a future oriented business . And, in our rapidly changing world, the fut... more Education is ‐ clearly ‐ a future oriented business . And, in our rapidly changing world, the future is shrouded in clouds of uncertai nty, and full of risks. Rather surprisingly, while reviewing the vast literature on topics relat ed to education, we noticed that the concept of risk is often missing from the discussions about social policies and individual options in education. And, on the rare occasions when the risk in education is still considered, the approaches adopted are based on concepts borrowed from financial theories that create risk instead of reducing it. This raises an ethical question for the educators: “to what extent are we ‐ policy makers, teachers, or other professionals in educati on - responsible for the professional path of our students 10 or 20 years from now, knowing that it is impossible to predict the social and economic evolution even for shorter time horizons?” In this paper we argue that the only means to provi de students with some robustness against t...
Predictive Toxicology (PT) attempts to describe the relationships between the chemical structure ... more Predictive Toxicology (PT) attempts to describe the relationships between the chemical structure of chemical compounds and biological and toxicological processes. The most important issue related to real-world PT problems is the huge number of the chemical descriptors. A secondary issue is the quality of the data since irrelevant, redundant, noisy, and unreliable data have a negative impact on the prediction results. The pre-processing step of Data Mining deals with complexity reduction as well as data quality improvement through feature selection, data cleaning, and noise reduction. In this paper, we present some of the issues that can be taken into account for preparing data before the actual knowledge discovery is performed.
In this paper, a possibilistic network implementation for uncertain knowledge modeling of the dia... more In this paper, a possibilistic network implementation for uncertain knowledge modeling of the diagnostic process is proposed as a means to achieve student diagnosis in intelligent tutoring system. This approach is proposed in the object oriented programming domain for diagnosis of students learning errors and misconception. In this expertise domain dependencies between data exist that are encoded in the structure of network. Also, it is available qualitative information about these data which are represented and interpreted with qualitative approach of possibility theory. The aim of student diagnosis system is to ensure an adapted support for the student and to sustain the student in personalized learning process and errors explanation.
Learning a Bayesian network from a numeric set of data is a challenging task because of dual natu... more Learning a Bayesian network from a numeric set of data is a challenging task because of dual nature of learning process: initial need to learn network structure, and then to find out the distribution probability tables. In this paper, we propose a machine-learning algorithm based on hill climbing search combined with Tabu list. The aim of learning process is to discover the best network that represents dependences between nodes. Another issue in machine learning procedure is handling numeric attributes. In order to do that, we must perform an attribute discretization pre-processes. This discretization operation can influence the results of learning network structure. Therefore, we make a comparative study to find out the most suitable combination between discretization method and learning algorithm, for a specific data set.
In this paper, a possibilistic network implementation for uncertain knowledge modeling of the dia... more In this paper, a possibilistic network implementation for uncertain knowledge modeling of the diagnostic process is proposed as a means to achieve student diagnosis in intelligent tutoring system. This approach is proposed in the object oriented programming domain for diagnosis of students learning errors and misconception. In this expertise domain dependencies between data exist that are encoded in the structure of network. Also, it is available qualitative information about these data which are represented and interpreted with qualitative approach of possibility theory. The aim of student diagnosis system is to ensure an adapted support for the student and to sustain the student in personalized learning process and errors explanation.
Many areas of artificial intelligence must handling with imperfection of information. One of the ... more Many areas of artificial intelligence must handling with imperfection of information. One of the ways to do this is using representation and reasoning with Bayesian networks. Creation of a Bayesian network consists in two stages. First stage is to design the node structure and directed links between them. Choosing of a structure for network can be done either through empirical developing by human experts or through machine learning algorithm. The second stage is completion of probability tables for each node. Using a machine learning method is useful, especially when we have a big amount of leaning data. But in many fields the amount of data is small, incomplete and inconsistent. In this paper, we make a case study for choosing the best learning method for small amount of learning data. Means more experiments we drop conclusion of using existent methods for learning a network structure.
2008 4th International IEEE Conference Intelligent Systems, 2008
... 535-13/09/2006. L. Dumitriu, C. Segal, MV. Craciun, A. Cocu are with the Computer Science Dep... more ... 535-13/09/2006. L. Dumitriu, C. Segal, MV. Craciun, A. Cocu are with the Computer Science Department, at Dunarea de Jos University, Galati, 800201 Romania (telephone: +40-236-460182, e-mail: {Luminita.Dumitriu, Cristina.Segal, Marian.Craciun, Adina.Cocu} @ ugal.ro). ...
This paper describes an experiment aimed to evaluate the influence of the environment in an "... more This paper describes an experiment aimed to evaluate the influence of the environment in an "iLab" (Innovation Laboratory) on the divergent thinking of a group of students. The philosophy behind the concept of iLab starts from the assumption that the convergent influence of an "extraordinary environment", a special communication application, and appropriate moderation techniques is capable to foster the creativity of the participants, and to enable a certain "collective intelligence". In our experiment, the group of 20 students was assigned the task to find as many new ideas as possible of ICT solutions for improving the quality of life of people with various disabilities. The ideas produced during the session were recorded and analyzed in terms of originality, feasibility, and utility using the FCA (Formal Concept Analysis) methodology. This paper presents the results of the experiment.
— There are several approaches in trying to solve the Quantitative 1Structure-Activity Relationsh... more — There are several approaches in trying to solve the Quantitative 1Structure-Activity Relationship (QSAR) problem. These approaches are based either on statistical methods or on predictive data mining. Among the statistical methods, one should consider regression analysis, pattern recognition (such as cluster analysis, factor analysis and principal components analysis) or partial least squares. Predictive data mining techniques use either neural networks, or genetic programming, or neuro-fuzzy knowledge. These approaches have a low explanatory capability or non at all. This paper attempts to establish a new approach in solving QSAR problems using descriptive data mining. This way, the relationship between the chemical properties and the activity of a substance would be comprehensibly modeled.
Applied Sciences
As the world population is ageing rapidly and old age comes with multiple health issues, the need... more As the world population is ageing rapidly and old age comes with multiple health issues, the need for medical services is likely to increase in a couple of decades beyond the limits of the medical systems of almost any country. In response to this trend, a variety of technologies have been developed with the aim of helping older people live independently as long as possible and reduce the burden on caregivers. In this paper, we propose a solution to encode the information about the activity of the monitored person, captured by a set of binary sensors, in the form of activity maps that reflect not only the intensity, but also the spatial distribution of the activity between a set of behaviorally meaningful places. Then, we propose a method for automatic analysis of the activity maps in order to detect deviations from the previously recorded routine. We have tested the method on two public activity recognition datasets and found that the proposed solution is not only feasible, but als...
The purpose of the paper was to develop an application that is capable to upload a picture and an... more The purpose of the paper was to develop an application that is capable to upload a picture and analyze it in order to determine melanoma lesions using artificial intelligence techniques. The proposed application is designed to use a previously trained convolutional neural network to recognize melanoma. For training, the examples from two known benchmarks were used and several attempts were made to find the best model driven by the neural network. The predictability rate is 0.95. The average time for obtaining the respond is 7 seconds.
Intelligent educational systems are knowledge-based systems (KBS) they can be developed by a gene... more Intelligent educational systems are knowledge-based systems (KBS) they can be developed by a generic knowledge-based system development methodology. In this paper, we present an ontology-based approach for formalizing different knowledge types. The formalism is based upon conceptual graphs. A priority concern to all research work in adaptive education is that of finding an appropriate representation for pedagogical knowledge. For implementation, we use the CoGITaNT environment (Conceptual Graphs Integrated Tools allowing Nested Typed graphs), a library of C++ classes (open-sources, developed by LIRMM CNRS, France) allowing the development of applications based on the CG knowledge representation scheme.
2020 IEEE Global Engineering Education Conference (EDUCON), 2020
This paper is a presentation of an educational project aimed at developing methodologies and curr... more This paper is a presentation of an educational project aimed at developing methodologies and curricula for teaching transversal skills and competences to engineering students. The project included a prior survey to identify needs and requirements of companies regarding the transversal skills that should possess the young adults who are about to enter the working world. The analysis of the survey’s responses required tailored research approaches to reveal correlations between the collected information. The research findings eventually stood as a basis to development of a project’s key component, the support methodology for teaching transversal-skills to students. The methodology included description of the learning goals, of means and activities needed to achieve these goals, as well as of methods and tools to evaluate the learning experiences. The pilot experimentation of the training courses provided valuable feedbacks for improving the educational approach.
2017 IEEE Global Engineering Education Conference (EDUCON), 2017
The students' creativity and “integrative thinking” are highly valued qualities in the modern... more The students' creativity and “integrative thinking” are highly valued qualities in the modern education. But is there any connection between them? This paper describes a simple experiment aimed to explore the possible connection and mutual influence between the students' creativity and their ability to summarize the main ideas from a given text. The overall creativity of the students who participated in the experiment was assessed by means of a previoulsly developped dedicated software tool. Besides that, their summary writing skills were assessed by a panel of 3 experts, and the scores were normalized to the interval [0-100]. Preliminary experimental data show a good correlation between the creativity quotient CQ and the ability of the students to extract and communicate the main ideas from a written text. Further research is in progress.
ICERI proceedings, Nov 1, 2022
Lecture notes in networks and systems, 2023
THE ANNALS OF “DUNAREA DE JOS” UNIVERSITY OF GALATI FASCICLE III, 2017, VOL. 40, NO. 1, ISSN 2344... more THE ANNALS OF “DUNAREA DE JOS” UNIVERSITY OF GALATI FASCICLE III, 2017, VOL. 40, NO. 1, ISSN 2344-4738, ISSN-L 1221-454X ELECTROTECHNICS, ELECTRONICS, AUTOMATIC CONTROL, INFORMATICS
The European Proceedings of Social and Behavioural Sciences, 2019
This is an Open Access article distributed under the terms of the Creative Commons Attribution-No... more This is an Open Access article distributed under the terms of the Creative Commons Attribution-Noncommercial 4.0 Unported License, permitting all non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
There are several approaches in trying to solve the Quantitative 1Structure-Activity Relationship... more There are several approaches in trying to solve the Quantitative 1Structure-Activity Relationship (QSAR) problem. These approaches are based either on statistical methods or on predictive data mining. Among the statistical methods, one should consider regression analysis, pattern recognition (such as cluster analysis, factor analysis and principal components analysis) or partial least squares. Predictive data mining techniques use either neural networks, or genetic programming, or neuro-fuzzy knowledge. These approaches have a low explanatory capability or non at all. This paper attempts to establish a new approach in solving QSAR problems using descriptive data mining. This way, the relationship between the chemical properties and the activity of a substance would be comprehensibly modeled.
Abstract: In intelligent tutoring systems, student or user modeling implies dealing with imperfec... more Abstract: In intelligent tutoring systems, student or user modeling implies dealing with imperfect and uncertain knowledge. One of the artificial intelligence techniques used for uncertainty management is that of Bayesian networks. This paradigm is recommended in the situation when exist dependencies between data and qualitative information about these data. In this work we present a student knowledge diagnosis model based on representation with Bayesian networks. The educational system incorporate a multimedia interface for accomplishes the testing tools. The results of testing sessions are represented and interpreted with probability theory in order to ensure an adapted support for the student. The aims of the computer assisted application that contains this diagnose module are to support the student in personalized learning process and errors explanation.
Education is ‐ clearly ‐ a future oriented business . And, in our rapidly changing world, the fut... more Education is ‐ clearly ‐ a future oriented business . And, in our rapidly changing world, the future is shrouded in clouds of uncertai nty, and full of risks. Rather surprisingly, while reviewing the vast literature on topics relat ed to education, we noticed that the concept of risk is often missing from the discussions about social policies and individual options in education. And, on the rare occasions when the risk in education is still considered, the approaches adopted are based on concepts borrowed from financial theories that create risk instead of reducing it. This raises an ethical question for the educators: “to what extent are we ‐ policy makers, teachers, or other professionals in educati on - responsible for the professional path of our students 10 or 20 years from now, knowing that it is impossible to predict the social and economic evolution even for shorter time horizons?” In this paper we argue that the only means to provi de students with some robustness against t...
Predictive Toxicology (PT) attempts to describe the relationships between the chemical structure ... more Predictive Toxicology (PT) attempts to describe the relationships between the chemical structure of chemical compounds and biological and toxicological processes. The most important issue related to real-world PT problems is the huge number of the chemical descriptors. A secondary issue is the quality of the data since irrelevant, redundant, noisy, and unreliable data have a negative impact on the prediction results. The pre-processing step of Data Mining deals with complexity reduction as well as data quality improvement through feature selection, data cleaning, and noise reduction. In this paper, we present some of the issues that can be taken into account for preparing data before the actual knowledge discovery is performed.
In this paper, a possibilistic network implementation for uncertain knowledge modeling of the dia... more In this paper, a possibilistic network implementation for uncertain knowledge modeling of the diagnostic process is proposed as a means to achieve student diagnosis in intelligent tutoring system. This approach is proposed in the object oriented programming domain for diagnosis of students learning errors and misconception. In this expertise domain dependencies between data exist that are encoded in the structure of network. Also, it is available qualitative information about these data which are represented and interpreted with qualitative approach of possibility theory. The aim of student diagnosis system is to ensure an adapted support for the student and to sustain the student in personalized learning process and errors explanation.
Learning a Bayesian network from a numeric set of data is a challenging task because of dual natu... more Learning a Bayesian network from a numeric set of data is a challenging task because of dual nature of learning process: initial need to learn network structure, and then to find out the distribution probability tables. In this paper, we propose a machine-learning algorithm based on hill climbing search combined with Tabu list. The aim of learning process is to discover the best network that represents dependences between nodes. Another issue in machine learning procedure is handling numeric attributes. In order to do that, we must perform an attribute discretization pre-processes. This discretization operation can influence the results of learning network structure. Therefore, we make a comparative study to find out the most suitable combination between discretization method and learning algorithm, for a specific data set.
In this paper, a possibilistic network implementation for uncertain knowledge modeling of the dia... more In this paper, a possibilistic network implementation for uncertain knowledge modeling of the diagnostic process is proposed as a means to achieve student diagnosis in intelligent tutoring system. This approach is proposed in the object oriented programming domain for diagnosis of students learning errors and misconception. In this expertise domain dependencies between data exist that are encoded in the structure of network. Also, it is available qualitative information about these data which are represented and interpreted with qualitative approach of possibility theory. The aim of student diagnosis system is to ensure an adapted support for the student and to sustain the student in personalized learning process and errors explanation.
Many areas of artificial intelligence must handling with imperfection of information. One of the ... more Many areas of artificial intelligence must handling with imperfection of information. One of the ways to do this is using representation and reasoning with Bayesian networks. Creation of a Bayesian network consists in two stages. First stage is to design the node structure and directed links between them. Choosing of a structure for network can be done either through empirical developing by human experts or through machine learning algorithm. The second stage is completion of probability tables for each node. Using a machine learning method is useful, especially when we have a big amount of leaning data. But in many fields the amount of data is small, incomplete and inconsistent. In this paper, we make a case study for choosing the best learning method for small amount of learning data. Means more experiments we drop conclusion of using existent methods for learning a network structure.
2008 4th International IEEE Conference Intelligent Systems, 2008
... 535-13/09/2006. L. Dumitriu, C. Segal, MV. Craciun, A. Cocu are with the Computer Science Dep... more ... 535-13/09/2006. L. Dumitriu, C. Segal, MV. Craciun, A. Cocu are with the Computer Science Department, at Dunarea de Jos University, Galati, 800201 Romania (telephone: +40-236-460182, e-mail: {Luminita.Dumitriu, Cristina.Segal, Marian.Craciun, Adina.Cocu} @ ugal.ro). ...
This paper describes an experiment aimed to evaluate the influence of the environment in an "... more This paper describes an experiment aimed to evaluate the influence of the environment in an "iLab" (Innovation Laboratory) on the divergent thinking of a group of students. The philosophy behind the concept of iLab starts from the assumption that the convergent influence of an "extraordinary environment", a special communication application, and appropriate moderation techniques is capable to foster the creativity of the participants, and to enable a certain "collective intelligence". In our experiment, the group of 20 students was assigned the task to find as many new ideas as possible of ICT solutions for improving the quality of life of people with various disabilities. The ideas produced during the session were recorded and analyzed in terms of originality, feasibility, and utility using the FCA (Formal Concept Analysis) methodology. This paper presents the results of the experiment.
— There are several approaches in trying to solve the Quantitative 1Structure-Activity Relationsh... more — There are several approaches in trying to solve the Quantitative 1Structure-Activity Relationship (QSAR) problem. These approaches are based either on statistical methods or on predictive data mining. Among the statistical methods, one should consider regression analysis, pattern recognition (such as cluster analysis, factor analysis and principal components analysis) or partial least squares. Predictive data mining techniques use either neural networks, or genetic programming, or neuro-fuzzy knowledge. These approaches have a low explanatory capability or non at all. This paper attempts to establish a new approach in solving QSAR problems using descriptive data mining. This way, the relationship between the chemical properties and the activity of a substance would be comprehensibly modeled.
Applied Sciences
As the world population is ageing rapidly and old age comes with multiple health issues, the need... more As the world population is ageing rapidly and old age comes with multiple health issues, the need for medical services is likely to increase in a couple of decades beyond the limits of the medical systems of almost any country. In response to this trend, a variety of technologies have been developed with the aim of helping older people live independently as long as possible and reduce the burden on caregivers. In this paper, we propose a solution to encode the information about the activity of the monitored person, captured by a set of binary sensors, in the form of activity maps that reflect not only the intensity, but also the spatial distribution of the activity between a set of behaviorally meaningful places. Then, we propose a method for automatic analysis of the activity maps in order to detect deviations from the previously recorded routine. We have tested the method on two public activity recognition datasets and found that the proposed solution is not only feasible, but als...
The purpose of the paper was to develop an application that is capable to upload a picture and an... more The purpose of the paper was to develop an application that is capable to upload a picture and analyze it in order to determine melanoma lesions using artificial intelligence techniques. The proposed application is designed to use a previously trained convolutional neural network to recognize melanoma. For training, the examples from two known benchmarks were used and several attempts were made to find the best model driven by the neural network. The predictability rate is 0.95. The average time for obtaining the respond is 7 seconds.
Intelligent educational systems are knowledge-based systems (KBS) they can be developed by a gene... more Intelligent educational systems are knowledge-based systems (KBS) they can be developed by a generic knowledge-based system development methodology. In this paper, we present an ontology-based approach for formalizing different knowledge types. The formalism is based upon conceptual graphs. A priority concern to all research work in adaptive education is that of finding an appropriate representation for pedagogical knowledge. For implementation, we use the CoGITaNT environment (Conceptual Graphs Integrated Tools allowing Nested Typed graphs), a library of C++ classes (open-sources, developed by LIRMM CNRS, France) allowing the development of applications based on the CG knowledge representation scheme.
2020 IEEE Global Engineering Education Conference (EDUCON), 2020
This paper is a presentation of an educational project aimed at developing methodologies and curr... more This paper is a presentation of an educational project aimed at developing methodologies and curricula for teaching transversal skills and competences to engineering students. The project included a prior survey to identify needs and requirements of companies regarding the transversal skills that should possess the young adults who are about to enter the working world. The analysis of the survey’s responses required tailored research approaches to reveal correlations between the collected information. The research findings eventually stood as a basis to development of a project’s key component, the support methodology for teaching transversal-skills to students. The methodology included description of the learning goals, of means and activities needed to achieve these goals, as well as of methods and tools to evaluate the learning experiences. The pilot experimentation of the training courses provided valuable feedbacks for improving the educational approach.
2017 IEEE Global Engineering Education Conference (EDUCON), 2017
The students' creativity and “integrative thinking” are highly valued qualities in the modern... more The students' creativity and “integrative thinking” are highly valued qualities in the modern education. But is there any connection between them? This paper describes a simple experiment aimed to explore the possible connection and mutual influence between the students' creativity and their ability to summarize the main ideas from a given text. The overall creativity of the students who participated in the experiment was assessed by means of a previoulsly developped dedicated software tool. Besides that, their summary writing skills were assessed by a panel of 3 experts, and the scores were normalized to the interval [0-100]. Preliminary experimental data show a good correlation between the creativity quotient CQ and the ability of the students to extract and communicate the main ideas from a written text. Further research is in progress.