Computer Sciences

Currently mobile games for educational purposes are a rapidly developing area. In this work we focus on multiplayer mobile games and its specific aspects such as communication and collaboration between players and learners. Therefore, it is important to understand how the communication and collaboration takes place in this type of games. This knowledge is valuable not only for further game development with an educational purpose (serious games) but in any type of mobile games. However, during our search we have observed that there is not enough analysis done in this topic. The first step, towards this research, was a literature overview of several authors' implementations of different communication types between players in educational mobile games. We analyzed those papers in the view of collaboration support. Also, this overview showed that some types of media are almost not used or not used for this purpose. Based on the outcomes of this literature overview, we decided to implement an educational mobile game, which promotes the collaboration in its core. Hence, choosing the technology for the development we put special attention to the types of communication that might support the collaboration in a specific game design. Furthermore, the proper decision on the communication types should support and keep in harmony the game-play in the educational context. This paper presents the concept of a multiplayer mobile game. As a technology for the player's communication and collaboration we have selected the one which are often used by mobile users but mostly not supported in mobile games-photo exchange. We call our game "PiX (Picture eXchange)". PiX is designed for attracting players to communication and collaboration by increasing their interest in an other persons' vision of the environment. The game enables players to see how different people interpret similar objects of the surroundings, what attracts people attention, and what kind of emotions they have.

Technologies are neither developed nor used in a sociocultural vacuum. They always embody and make use of some kind of tacit knowledge. The gulf between the knowledge that designers possess and the knowledge that users have increases when ICTs are distributed in non-Western developing countries. If an ICT project is ignorant about how to effect local implementation, progress will probably be negligible, and if the planning of an ICT project is irrelevant to local needs and conditions, what may seem like progress may actually not improve anything at all. In this paper we outline our ICT education development project for a Tanzanian college, introduce a model for analysing ICT projects in cultural contexts, and analyse our Tanzanian project in terms of that model. We argue that successful contextualisation takes cognizance of at least four levels (or phases) of projects. These are import, transfer, apply, and contextualise.

This study analyzes the possibility of involving Computer Science students in multiplayer mobile game development processes and using their knowledge to solve collaboration problems. We describe the importance of Computer Science as one of the Computing disciplines, and provide a summary of its knowledge areas according to their ability to help solve collaboration problems. This analysis is based on the Bachelor's degree study program at the University of Joensuu, Finland. Research shows that it is highly possible to involve students from their first year of study in solving collaboration problems of multiplayer mobile games.

In many parts of Africa the application of information and communication technologies (ICTs) in higher education institutions offers a particularly interesting research problem. This is because of the far-reaching consequences that information and communication technology (ICT) services and education have on the university graduates during their careers in African societies. Our extensive and well-documented experience of utilizing ICTs in a Tanzanian private university has lead us to propose a four-level approach, called CATI model, that can be used as the basis for implementing new ICTs, for evaluating how ICT is used, and for planning ICT education in developing countries. We conceptually categorize the four levels that we use in our model – contextualize, apply, transfer and import – as an ethnocomputing approach. This four-level model enable educators to use ICT services and ICT education in African universities in a way that will prepare university graduates to make a positive...

Technologies are neither developed nor used in a sociocultural vacuum. They always embody and make use of some kind of tacit knowledge. The gulf between the knowledge that designers possess and the knowledge that users have increases when ICTs are distributed in non-Western developing countries. If an ICT project is ignorant about how to effect local implementation, progress will probably be negligible, and if the planning of an ICT project is irrelevant to local needs and conditions, what may seem like progress may actually not improve anything at all. In this paper we outline our ICT education development project for a Tanzanian college, introduce a model for analysing ICT projects in cultural contexts, and analyse our Tanzanian project in terms of that model. We argue that successful contextualisation takes cognizance of at least four levels (or phases) of projects. These are import, transfer, apply, and contextualise.

Digital games have traditionally been targeted at younger generations, although the proportion of older adult players is increasing. However, the design processes of digital games often do not consider the special needs of older adults. Co-design is a potential method to address this, but there is little research on co-designing games with older adults. In our study, we proposed a co-design process model that considers the intergenerational perspective. Using this model, eight older adults (two males and six females aged 47–80) and 22 sixth graders (11 males and 11 females aged 12–13) co-designed a digital game. The content of the game was based on old concepts used by the designers during their childhood. Similarly, game content involving new words and concepts were produced by the sixth graders. We collected data using semi-structured interviews and observations during the co-design process over a period of 24 months and then processed the data using grounded theory. The results i...

, and rest of the Hannover group for their kind hospitality during my visit in Hannover. I also want to thank my parents, Paavo and Riitta, and my sister Johanna for all the support. Finally I express my deep gratitude and admiration to Anu Saarinen. She has brought meaning into my life and has made this work worth doing.

Tutkielma keskittyy käsittelemään ohjelmistovirheiden luokitteluiden hyödyntämistä
ohjelmistotuotantoprosessin kehittämisessä. Tutkielmassa käsitellään yleisesti kehittämisen tärkeyttä, kehittämismallia, virheluokkia ja kehittyneempää SAWO virheluokittelua. Luokitteluiden käyttöä yrityksissä pyritään selvittämään suoritetulla kyselytutkimuksella. Kyselytutkimus pyrkii selvittämään ja saamaan näkemystä luokitteluiden käytöstä ohjelmistotuotantoprosessin kehittämisessä suomalaisissa yrityksissä

Tietojärjestelmillä on erittäin suuri rooli ja merkitys yritysten jokapäiväisessä toiminnassa. Näin ollen tietojärjestelmien häiriötön toiminta sekä tietojärjestelmäprojektien
tehokas toteuttaminen ovat yrityksen toiminnan kannalta ensisijaisen tärkeitä. Tehokas
vianhallinta edistää näitä tavoitteita.
Virheluokittelukäytännöt kehittyvät ajan edetessä tarkemmiksi ja tehokkaammiksi,
niillä voidaan entistä paremmin hallinnoida ja analysoida havaittuja virheitä projekteissa sekä kehittää ohjelmistokehitysprosessia entistä paremmaksi. Lisäksi varsinaiset
kehittämisen elinkaarimallit ovat kehittyneet vesiputousmallista ketteriin malleihin
(Agile) ja DevOps-malleihin, jotka ovat entistä tehokkaampia ja vähemmän virheitä
tuottavia käytäntöjä. Elinkaarimallit ovat kehittymässä suuntaan, jossa mallit itsessään
vähentävät virheitä lisääntyvän automaation avulla. Tutkittavaksi asiaksi nousee virheluokitteluiden merkitys nykyaikaisissa kehittämismalleissa.
Tässä tutkielmassa vikojen hallintaa tarkastellaan esittelemällä erilaisia virheiden luokittelutapoja/-menetelmiä. Näiden lisäksi tutkielmassa on kuvattu yksinkertainen, tehokas ja helposti toteutettava ohjelmistokehitysprosessin kehittämismalli, joka pohjautuu virheiden luokitteluun ja analysointiin. Projektien tehokasta läpivientiä tarkastellaan esittelemällä sovelluskehityksen erilaisia elinkaarimalleja.
Tutkielmassa käsitellään virheiden määritelmiä, virheluokitteluiden kuvauksia, moderneja elinkaarimalleja kuten agile- ja DevOps-mallit sekä uusia kehittämisen termejä
kuten kehitysputki (pipeline) ja konttikehitys (container). Tutkielmassa käsitellään lyhyesti myös perinteistä vesiputousmallia. Tutkielma keskittyy käsittelemään pääasiassa modernien elinkaarimallien sekä kehitysputki ja konttikehitys -termien suhdetta
virheluokitteluihin ja selvittämään virheluokitteluiden tarpeellisuutta ja hyödyntämistä
näissä elinkaarimalleissa.
Tutkielmassa on käytetty tieteellisiä tietolähteitä, toteutettu tietokantahakuja mainituista kehitysmenetelmistä, käsitelty hakutuloksia, tehty yhteenvetoa ja pohdinta

One of the difficulties in understanding and debugging spreadsheets is due to the invisibility of the data flow structure which is associated with cell formulas. In this paper, we present a spreadsheet visualization approach that is mainly based on the Markov Clustering (MCL) algorithm in an attempt to help spreadsheet users understand and debug their spreadsheets.

Spreadsheets are computer programs. However, as with programs written in traditional programming languages, it is very difficult to understand a spreadsheet created by others. This is because spreadsheet users normally view the superficial numerical (value) view of spreadsheets although computations are specified through "hidden" cell formulas. The cell formulas also define the data-flow structure of the spreadsheet. In this paper, we present a technique that highlights logical areas in spreadsheets which may act as a guide in understanding a spreadsheet. Instead of focussing their attention on the whole spreadsheet, spreadsheet users may narrow their focus to one logical area at a time. We identify logical areas in spreadsheets by using the MCL (Markov Clustering) algorithm on the underlying spreadsheet data-flow graph.

Earlier research on spreadsheet authors' mental models has shown that people possess several types of information about spreadsheets. Especially, when explaining a spreadsheet, the real-world and domain mental models are prominent and the spreadsheet model is suppressed, but when locating and fixing an error, one must constantly switch back and forth between the domain model and the spreadsheet model, which requires frequent use of the mapping between problem domain concepts and their spreadsheet model counterparts. This paper introduces a new spreadsheet visualization tool and its empirical evaluation. The tool translates traditional spreadsheet formulas into problem domain narratives and highlights referenced cells. The tool was found to be easy to learn and helped the participants to locate more errors in spreadsheets. Furthermore, the tool increased the use of the domain mental model and appeared to improve the mapping between the spreadsheet model and the domain model.

This thesis tackled the problem of errors in spreadsheets by studying mental models of spreadsheet authors performing various spreadsheet activities. Using a constructive research approach, we conducted three empirical studies and developed two spreadsheet visualization tools. Our results show the need for spreadsheet tools that reflect spreadsheet authors’ mental models, and for a paradigm shift which allows spreadsheets to be developed in terms of a mental model of the application domain.

In this paper, we explore the potential of introducing fundamental concepts of programming to first year students using spreadsheets. Many students find programming to be difficult to learn as the concepts are new and unfamiliar. On the other hand, spreadsheets have a relatively easy to use interface that provides instant viewing of results of data manipulation. Fundamental concepts of programming can then be indirectly introduced using their spreadsheet equivalents before the students start their formal introductory programming class. This paper specifically focusses on how fundamental programming concepts can be mapped to their spreadsheet equivalents.

Most errors in spreadsheets are formula-based. Referenced cells in formulas are normally presented using the traditional A1 cell referencing style. A spreadsheet user has to therefore mentally map referenced
cells to their corresponding labels in order to comprehend a formula in the context of the problem domain. In this paper, we give a detailed description of an algorithm that can be used to dynamically translate traditional spreadsheet formulas to their problem domain equivalents which are easier to understand. The translation is done as one accesses a formula cell in a spreadsheet. The formula translation is based on inferred labels of referenced cells in the formula. The aim of the translation is to ease the cognitive load on the spreadsheet user and hence improving the error-prone spreadsheet development process. The paper also highlights some factors that need to be taken into consideration when dynamically translating spreadsheet formulas. The number of referenced cells in a formula and the distance of labels from referenced cells will determine the speed of the translation process hence affecting system responsiveness as one navigates through a spreadsheet. Unpredictable spatial arrangement of data in spreadsheets (spreadsheet layout) can also pose a challenge to the translation algorithm which may lead to mis-translation of spreadsheet formulas. These challenges might increase the cognitive load on the spreadsheet user hence negating the purpose of dynamically translating spreadsheet formulas.

In this thesis, we tackled the problem of errors in spreadsheets by studying spreadsheet authors' mental models. It is a common assertion that humans have mental models of the systems they interact with, and it is difficult to explain many aspects of human behaviour without resorting to a construct such as mental models. We therefore argue that it is important to first of all understand what types of mental models spreadsheet authors possess when they are doing different spreadsheet tasks in order to better understand why the spreadsheet process is so error-prone and to be able to devise new tools that better correspond to the way they think. I thank the Almighty God for the gift of life and for the grace to undertake this research work. My profound thanks should also go to Prof. Jorma Sajaniemi, my supervisor. I can not find words enough to express my appreciation for his wonderful supervisory skills. Indeed, I feel honoured to learn under his tutelage. The memories of the past four years, I shall always cherish. Dr. Marja Kuittinen should also receive my special vote of thanks for her help in some aspects of the studies in the course of this work, and particularly, for her help in the validation of the data coding method in the first empirical study conducted in this research work. I also sincerely thank the preliminary examiners of this dissertation, Prof.

To address the problem of errors in spreadsheets, we have investigated spreadsheet authors' mental models in a hope of finding cognition-based principles for spreadsheet visualization and debugging tools. To this end, we have conducted three empirical studies. The first study explored the nature of mental models of spreadsheet authors during explaining and debugging tasks. It was found that several mental models about spreadsheets are activated in spreadsheet authors' minds. Particularly, when explaining a spreadsheet, the real-world and domain mental models are prominent, and the spreadsheet model is suppressed; however, when locating and fixing an error, one must constantly switch back and forth between the domain model and the spreadsheet model, which requires frequent use of the mapping between problem domain concepts and their spreadsheet model counterparts. The second study examined the effects of replacing traditional spreadsheet formulas with problem domain narratives in the context of a debugging task. Domain narratives were found to be easy to learn and they helped participants to locate more errors in spreadsheets. Furthermore, domain narratives also increased the use of the domain mental model and appeared to improve the mapping between the domain and spreadsheet models. The third study investigated the effects of allowing spreadsheet authors to fix errors by editing domain narratives, thus relieving them from the use of traditional low-level cell references. This scenario was found to promote spreadsheet authors using even more of their domain mental model in a manner that completely overshadowed the use of their spreadsheet mental model. Thus, from a mental model perspective, it is possible to devise a new spreadsheet paradigm that uses domain narratives in place of traditional spreadsheet formulas, thus automatically presenting spreadsheet content so that it prompts spreadsheet authors to think in a manner that closely corresponds to their mental models of the application domain.

The increasing volume of customs declarations presents a significant challenge for customs administrations in many countries, including
Sub-Saharan African countries. Manual inspection of customs declarations is particularly difficult due to limited human and technical
resources as well as limited knowledge about traded products by customs
officials. These factors have made the detection of customs fraud
unreliable, resulting in lost tax revenue to governments, among many
other effects. In this study, we address this challenge by using a machine
learning-based solution to detect customs fraud. We applied the design
science research methodology to develop and compare the predictive performance of Support Vector Machine (SVM) and Random Forest (RF)
models in detecting misclassification fraud, also known as Harmonised
System (HS) fraud. We utilised transaction-level customs data from the
Malawi Revenue Authority, from a developing country in Sub-Saharan
Africa (SSA). Our study’s findings suggest that the RF model outperformed
the SVM model in all key classification evaluation metrics. On the one hand, the model achieved an accuracy score of 0.958, a precision
score of 0.919, a recall score of 0.850, a precision-recall area under the
curve (PR-AUC) score of 0.953, and an F1 score of 0.883. On the other
hand, the SVM model produced an accuracy score of 0.925, a precision
score of 0.870, a recall score of 0.700, a precision-recall area under the
curve (PR-AUC) score of 0.872, and an F1 score of 0.776. This study
makes three main contributions: Firstly, the study used a dataset from
a developing country, unlike other studies that use datasets from developed countries. Secondly, the study employed transaction-level customs import data collected at an entire 8-digit HS code level, while most similar studies used HS codes at 2, 4, or 6-digit levels, which provide more general descriptions of the traded products. Lastly, unlike similar studies, this study specifically compared the capabilities of the SVM and RF algorithms in detecting customs fraud.