Summary: The Internet of Things has been gaining prominence and importance around the world. Its application allows, for example, an optimized use of resources in the environment where it is applied. For this reason, the National Development Bank has proposed an action plan for the application of the Internet of Things in Brazil. The plan contemplates four distinct environments (cities, health, rural and industries), each with its specific strategic objectives. In line with this action plan, this work presents a solution based on the Internet of Things for the rural environment, more specifically, for a plantation of grapes for the production of wines, focusing on the strategic objective of the efficient use of natural resources and inputs. The solution includes the use of sensors and actuators networks, aiming at efficiency in irrigation, prevention and action in fires, detection and combat of plantation diseases, in addition to data collection and analysis, aiming at guaranteeing the quality of future crops.
Abstract: A Concept Inventory (CI) is a set of multiple-choice questions that can be used to assess the students' comprehension on some topic at some point during a course. Each incorrect choice corresponds to a specific misconception - an inaccurate line of thought students often follows. In previous work we identified misconceptions for Introductory Programming Courses (CS1) in C and Python programming languages. Based on these misconceptions, in this work we mapped and documented possible misconceptions that could be found in CS1 courses based on JAVA programming language. We developed Java programming codes in order to check if the misconceptions detected in C and Python also could occur in Java language. Also, we used these codes to identify possible new misconceptions in Java, exploring its syntax and object-oriented paradigm. Finally, the misconceptions identified (N = 34) were classified into 8 programming topics and documented through an Antipattern structure, composed by: code (a label to identify the misconception); name; description (a brief explanation about the misconception); example; rationale (a possible reason why the misconception happens); consequences; detection (where and how the misconception happens); and improvement (solutions in order to prevent the misconception). Future work reports to the validation of the misconceptions identified on this work, through CS1 specialists and students, following the CI design process proposed by Almstrum et al.
Abstract: The purpose of this work is to document the development process of a Concept Inventory for introductory Computer Science in the Python programming language. It covers the analysis of our previous work in the C language, our inspection of past exam answers from two sections of a CS1 course in Python, and interviews with the instructors who gave these sections. From our previous list of 33 misconceptions in C, we have maintained 17, and raised new 11 hypotheses specific to the Python language, for a total of 28 hypothetical misconceptions to be validated as a further step of this study.
Abstract: Technology has become increasingly present in all spheres of society. However, the way we interact with technological devices has undergone significant changes. For example, in Internet of Things scenarios, functionality may be distributed on different devices with distinct capabilities; this distribution may cause interaction shifted in time and space. In this context, the human-computer interaction seeks to provide quality of interaction through measures coming from User Experience (UX) evaluations. This work presents an exploratory study using three UX evaluation methods: Attrakdiff, Emocards, and UX Curve, which include aspects such as temporality, emotion, hedonic and pragmatic characteristics. For the accomplishment of the study, a wearable device was used to emphasize the characteristics of IoT. The results showed gaps in the methods in the evaluation of UX in this scenario.
Abstract: The Conceptual Inventory project, developed by students, teachers and researchers at the Unicamp Institute of Computing, has as its initial proposal the identification of problems of understanding (misconceptions) of students of introductory programming courses, such as MC102. From this identification, Conceptual Inventories (CIs) were created, multiple-choice questionnaires in which each incorrect alternative is mapped to one of these misconceptions. One of the objectives of the project is to allow teachers from the most varied institutions in the world to, in an automated way, create, manage, submit to their students, and view reports of the application of CIs. Thus, one of the project's needs is the use of a tool that at the same time allows: a) the automation of the process of creating, managing and submitting questionnaires and; b) the possibility of customizing the data recorded in the system, such as the extraction of specific information and response patterns from the questionnaires answered. This Technical Report presents the implementation of a web tool that meets these needs.
Abstract: This Technical Report is part of an ongoing work to develop and assess a Concept Inventory (CI) for Introductory Computer Programming Courses (CS1) based on the C programming language. A CI is a set of multiple-choice questions that can be used to assess the students' comprehension on some topic at some point during a course. Each incorrect choice corresponds to a specific misconception - an inaccurate line of thought students often follow. CIs exist for several knowledge areas, such as physics, chemistry, and statistics. In previous work we identified, through the analysis of open-ended exams and interviews with instructors, 19 misconceptions related to CS1 courses in the C programming language. The misconceptions were grouped into 7 topics: function parameter use and scope; variables; recursion; iteration; structures; pointers; and boolean expressions. In this work, we: a) present in details the 19 misconceptions identified on the previous work and; b) present the Exploratory Questionnaire created to assess the previous work's misconceptions and also allow the identification of new misconceptions. This Technical Report will be referenced by an article that explains the Exploratory Questionnaire design and the results of its online administration, including the student's performance and individual think-aloud interviews with selected participants.
Summary: The Project OpenDesign aims to characterize and conceptualize an `open 'design process of interactive systems based on the Socially Conscious Design model. One of the objectives of the Project is the development and experimentation of a platform based on the proposed process model. The activities developed in the Project have moved towards this platform that offers the necessary tools for Open Design. This report presents the team's views on the platform that were collected in participatory activities developed to support conception and understanding of the concept of Open Design with interested parties. In order to deepen our understanding and guide development activities, we present, next, a usage scenario to characterize the platform OpenDesign proposal.
Summary: The OpenDesign project aims to support the design of computational solutions to problems of broad scope, in a systemic and socially conscious view. Development practices with universal access to artifacts, transparent design environment, peer review, informal communication, frequent feedback, design and development walking in parallel, are some aspects that characterize the project and the desired Platform. This technical report presents how the team has characterized and conducted its internal development process for the OpenDesign system, especially considering the goal of the Platform's development. To this end, it describes and discusses Process models (eg agile models), use of tools (eg Trello), and Project choices that culminated in a Process Model idealized for the OpenDesign project.
Summary: In order to understand a design problem, it is necessary for the team to organize itself for the study and discussions while documenting the activities developed and conclusions. The research project OpenDesign aims to characterize and formalize a design and development process, and to conceive and experience a support platform for the use of techniques and artifacts for the socially conscious design of computer systems inspired by philosophy open source. This report shows how the team used some artifacts of Socially Conscious Design available in the DSC tool to clarify and understand the problem of designing a platform for Open Design. As a result, the team identified stakeholders, demands, requirements and purposes of use for the platform, in addition to mapping the features of the DSC tool that support socially conscious design as the basis for the evolution of a platform OpenDesign.
Summary: The understanding and survey of the functionalities and quality attributes of a system play a central role in its acceptance and proper use. Non-conventional systems that involve uncertainties about their requirements demand methodologies that adequately support the capture of the needs, desires and objectives of the interested parties. In the context of the OpenDesign project, a platform that allows the collaborative elaboration of projects has been investigated. Understanding the desired features on this platform is a non-trivial activity. In this technical report, it is proposed to use user stories elucidated through participatory practices as an alternative to obtain a clarification of the platform's functionalities and their possible forms of use. This paper describes the definition, prioritization and refinement of user stories for the OpenDesign project. Based on these results, we defined an organization of the central concepts that permeate the OpenDesign proposal, which are expressed in a map with the synthesis of functionalities.
Abstract: The educational environment usually emphasizes learning as a declarative and procedural process. However, there are other crucial components that can be considered, such as the Perceptual Learning (PL), related to the ways the information is perceived and extracted. A way to model the PL approach is through Perceptual Learning Modules (PLMs), a set of multiple-choice questions in which each available choice is mapped to a learning pattern. Currently, there are initiatives related to PLM practices in mathematics, flight learning, medicine and language educational areas. This Technical Report is a reference document, not peer-reviewed, that aims to document a set of 64 PLM multiple-choice questions designed to Introductory Programming (CS1) courses in Python language, therefore allowing our work to be replicated by other researchers. The actual use of the PLM questions and the analysis and discussion of its impact on a CS1 course will be addressed in further works.
Summary: We started to restructure these notes in 2016, when Meidanis was on sabbatical at the University of Ottawa, at David Sankoff's laboratory. We use the term “restructure” because the main results displayed here, on minimax genomes under rank, first appeared in the text presented by Biller in his doctoral qualification exam, written in 2014. The content is divided into eight chapters, as follows. In Chapter 1 we introduced the first definitions, including genomic matrices, distance, and orbits. We also derived an important formula for distance based on orbits. In Chapter 2 we defined operations on genomes, with a focus on those in a small post. The chapter ends with the definition of basic operations, namely, cuts, joins, and double exchanges. In Chapter 3, we studied ordering scenarios going from one genome to another through basic operations. We show that from each genome we can reach any other with such scenarios. This also provides an alternative way of calculating the distance. Intermediate genomes are the topic of Chapter 4. They can be characterized as members of optimal scenarios, or as genomes for which there is equality in triangular inequality. They are also the medians of two genomes. A related notion is the notion of minimax genomes, explored in Chapter 5. We establish a lower limit for the minimax score, and show exactly the cases where it is possible to reach this limit. In any case, it is always possible to find a genome less than one unit from the lower limit. Chapter 6 deals with an interesting rank distance parity property. In Chapter 7, we tried to approximate mathematical definitions to the underlying biological concepts. Finally, Chapter 8 contains exercises on all the other chapters in the report, some of which have solutions or tips.
Abstract: We started the work of reshaping these notes in 2016, when Meidanis was on sabbatical at the University of Ottawa, in David Sankoff's lab. We use the term “reshaping” because the main results shown here, on minimax genomes under the rank distance, were first presented in Biller's text for her PhD qualifying exam, written in 2014. The contents are divided into eight chapters, as follows. In Chapter 1 we introduce the first definitions, including genome matrices, distance, and orbits. We also derive an important formula for the distance based on orbits. In Chapter 2 we define operations on genomes, with focus on those with small rank. The chapter closes with the definition of basic operations, namely, cuts, joins, and double swaps. In Chapter 3 we study sorting scenarios going from a genome to another by basic operations. We show that from every genome we can reach any other with such scenarios. This also provides an alternative way of computing the distance. Intermediate genomes are the topic of Chapter 4. They can be characterized both as optimal scenario members, and as genomes for which the triangle inequality becomes an equality. They are also the medians of two genomes. A related notion is that of a minimax genome, explored in Chapter 5. We establish a lower bound for the minimax score, and show exactly the cases where it is possible to achieve such a score. In any case, it is always possible to find a genome within 1 unit of the lower bound. Chapter 6 deals with an interesting parity property of the rank distance. Chapter 7 tried to bridge the gap between mathematical definition and the real biological concepts. Finally, Chapter 8 contains exercises on the entire contents of the report, some of them with solutions or hints.
Summary: Automatic identification of mappings between ontologies in different languages involves the study of similarity measures. These measures calculate the syntactic and / or semantic proximity between concepts of ontologies. In this work, Research developed with financial support from FAPESP, we conducted a series of experiments to evaluate approaches to similarity measures and developed a technique to compute the similarity between two concepts, based on the weighting between syntactic and semantic similarities. We used the NASARI vectors in conjunction with the BabelNet neutral domain semantic network to calculate semantic similarity. Results indicate that the weight assignment in the measures positively influences the quality of the obtained mappings. The effect of combining the measures results in better results than using each form of similarity separately.
Abstract: In previous work we implemented and compared three learning approaches in a Computer Science introductory programming course (CS1): the traditional lectured-based learning; the problem-based learning; and the Peer Instruction (PI). The study also pointed out guidelines to support a customized and more effective approach of the PI to the Computer Science environment, defined as CSPI (Computer Science Peer Instruction). In this work we present data related to a first CSPI use in some classes of a CS1 course taught in Python language (MC102). Specifically, this study presents the usability and reliability data relative to the use of clickers (student response systems) to support the CSPI approach. As the research is still on development, we plan to discuss the CSPI approach, its educational impact and assessment in future works.
Abstract: An important problem in genome comparison is the genome sorting problem, that is to find a sequence of basic operations that transforms one genome into another and corresponds to the distance between them. These sequences are called optimal sorting scenarios. However, there is usually a large number of such scenarios, and a naive algorithm is very likely to be biased towards a specific type of scenario, impairing its usefulness in real-world applications. One way to go beyond the traditional sorting algorithms is to explore all possible solutions, looking at all the optimal sorting scenarios instead of just an arbitrary one. Another approach in the same direction is to analyze all the intermediate genomes, that is, all genomes that are part of an optimal sorting scenario. In this paper, we show how to count the number of optimal sorting scenarios and the number of intermediate genomes between any two given genomes, under the rank distance.
Abstract: This work is a report related to the development and assessment of a Concept Inventory to Introductory Programming (CS1) Courses. A Concept Inventory (CI) is a set of multiple-choice questions addressing specific misunderstandings and misconceptions of the students. In previous works, through instructor interviews, exam analysis, online pilot test and interviews with students, we have identified a list of 33 misconceptions related to 7 programming topics in C language. On this report, we present a CI composed of 27 multiple-choice questions in C language. Each possible answer, besides the right one, was mapped to a previously documented misconception. Future work involves the CI submission to CS1 students and the analysis of its internal consistency and educational impact.
Abstract: Emotionally annotated corpora are especially important for training machine learning models for automatic emotion identification, among other applications. However, the task of manually assigning emotions to a corpus carries a high level of subjectivity. In this technical report, we describe the annotation tools and methodology we used for dealing with this challenge when building an emotionally annotated corpus of investor tweets.
Abstract: Computational ontologies refer to formal representation models that characterize domain concepts. These artifacts are relevant to shape the meanings of data in an explicit way for humans and machines. The study of socio-enactive systems demands the understanding of human actions and the adequate system response in a cyclical interaction. Ontologies can play a key role in the development of socio-enactive systems by providing meaning to data entering and leaving the systems. However, this research field still lacks thorough literature studies. This technical report presents the results achieved by a research working group to further develop methods for ontology engineering in the context of socio-enactive systems. We present a literature review, technical solutions, in addition to conceptual and practical outcomes. Our findings include a preliminary modeling of concepts involved in socio-enactive systems, and a software architecture that allows ontology-based interpretation of several types of data. This research has been developed in the socio-enactive systems project, FAPESP's thematic project (grant # 2015 / 16528-0).
Abstract: This technical report presents the preliminary results from the work group “GT Hospital”, a subgroup from the Socio-enactive Systems research group. GT Hospital will explore the “socio-enactive” concept within the context of a hospital for craniofacial rehabilitation. In this first year of project, this group has explored the scenario with the lens of Organizational Semiotics via DSC (Socially Aware Design) system; developed preliminary products for proof of concept and a communication protocol. This research is supported by FAPESP, process # 2015/16528.
Abstract: This technical report presents preliminary results regarding the museum research scenario developed during the first year of the Socioenactive Systems project. In this report, we first introduce the research scenario, with the research goals and methodology. Then, we present a preliminary literature review by looking at the intersections between the concepts of enactive systems, interactive art and universal design. Next, we show our preliminary results, which include two exploratory artifacts, as well as research projects and publications associated with the research scenario. Lastly, we ponder on how our efforts during this first year are aligned with the project goals, and we also present future steps for the following year.
Summary: This work is a partial report of the multiannual FAPESP's thematic project “Socio-Enactive Systems: Investigating New Dimensions in the Design of Interaction Mediated by Information and Communication Technologies”. Specifically, on this report we focus on the study, exploration and assessment of socio-enactive solutions in the educational environment. The methodology used was the bottom-up approach: initially we analyzed the related literature; then we created high-level socio-enactive work scenarios (eg inside or outside the classroom, considering K12 or undergraduate students, etc.); then we defined and expanded a more specific work scenario (children in the classroom environment). Finally we created a MVP (Minimum Viable Product) for the work scenario, developing and implementing a prototype through the MakeBlock platform (an arduino based robot vehicle), using Scratch as programming language. In the work scenario we propose the creation of an environment where children must use concepts of logic and computational thinking to position themselves on a stage in order to predict the robot's movement patterns. The goal is enable the kids, through their own movements, to guide the robot to a predefined location at the center of the stage. Proof of concept experiments indicate the need for prototype adjustments and adaptations. Next steps include the design and performance of workshops with the children, aiming at retrieving inloco information that will support further development of socio-enactive educational systems.
|
Instituto de Computação :: State University of Campinas
Av. Albert Einstein, 1251 - Cidade Universitária Zeferino Vaz • 13083-852 Campinas, SP - Brazil • Phone: [19] 3521-5838 |