Master of Science / Department of Computing and Information Sciences / William H. Hsu / In this thesis, I examine the Articles for Deletion (AfD) system in /Wikipedia/, a large-scale collaborative editing project. Articles in Wikipedia can be nominated for deletion by registered users, who are expected to cite criteria for deletion from the Wikipedia deletion. For example, an article can be nominated for deletion if there are any copyright violations, vandalism, advertising or other spam without relevant content, advertising or other spam without relevant content. Articles whose subject matter does not meet the notability criteria or any other content not suitable for an encyclopedia are also subject to deletion. The AfD page for an article is where Wikipedians (users of Wikipedia) discuss whether an article should be deleted. Articles listed are normally discussed for at least seven days, after which the deletion process proceeds based on community consensus. Then the page may be kept, merged or redirected, transwikied (i.e., copied to another Wikimedia project), renamed/moved to another title, userfied or migrated to a user subpage, or deleted per the deletion policy. Users can vote to keep, delete or merge the nominated article. These votes can be viewed in article’s view AfD page. However, this polling does not necessarily determine the outcome of the AfD process; in fact, Wikipedia policy specifically stipulates that a vote tally alone should not be considered sufficient basis for a decision to delete or retain a page. In this research, I apply machine learning methods to determine how the final outcome of an AfD process is affected by factors such as the difference between versions of an article, number of edits, and number of disjoint edits (according to some contiguity constraints). My goal is to predict the outcome of an AfD by analyzing the AfD page and editing history of the article. The technical objectives are to extract features from the AfD discussion and version history, as reflected in the edit history page, that reflect factors such as those discussed above, can be tested for relevance, and provide a basis for inductive generalization over past AfDs. Applications of such feature analysis include prediction and recommendation, with the performance goal of improving the precision and recall of AfD outcome prediction.

Development of early leafspot, caused by Cercospora arachidicola Hori, was monitored on' Florigiant' peanut (Arachis hypogaea L.) at two field sites in Suffolk, Virginia. In one study, plants in 27-cm-diameter plots were inoculated with 20,000 conidia and inoculation dates were replicated in five randomized complete blocks. At location one in 1985 and 1986, lesions/leaf at two weeks after inoculation correlated significantly (P ≤ 0.05) with infection indices (IND) developed by the Virginia leafs pot advisory and hours of relative humidity (RH) ≥ 95%. At location two, correlations between lesions/leaf and IND as well as hours of RH ≥ 95% were significant in 1986, but not in 1985. Certain site specific factors were believed to have altered plant susceptibility to leafspot at this site in 1985. In another study, pots with greenhouse-grown peanut were placed between unsprayed rows of field plants, heavily colonized by C. arachidicola. Plants were removed after 3, 5, and 7 days of field exposure for six consecutive weeks in 1986 and returned to the greenhouse. Lesions/leaf at two weeks after initial exposure were correlated with IND values computed by five versions of the leaf spot advisory. Significant correlations were found between lesions/leaf on plants with field exposures of 5 and 7 days and cumulative IND values and hours of RH ≥ 90% and 95%. The low incidence of lesions resulting with field exposures of only 3 days coupled with a lack of significant correlations between disease and cumulative IND values for 3 days after inoculation in both studies suggests that infection processes require several days, and that fungicides may be applied to achieve disease control during this time. / M.S.

Wikipedia is a free, web-based, collaborative, multilingual encyclopedia project supported by the non-profit Wikimedia Foundation. Due to the free nature of Wikipedia and allowing open access to everyone to edit articles the quality of articles may be affected. As all people don’t have equal level of knowledge and also different people have different opinions about a topic so there may be difference between the contributions made by different authors. To overcome this situation it is very important to classify the articles so that the articles of good quality can be separated from the poor quality articles and should be removed from the database. The aim of this study is to classify the articles of Wikipedia into two classes class 0 (poor quality) and class 1(good quality) using the Adaptive Neuro Fuzzy Inference System (ANFIS) and data mining techniques. Two ANFIS are built using the Fuzzy Logic Toolbox [1] available in Matlab. The first ANFIS is based on the rules obtained from J48 classifier in WEKA while the other one was built by using the expert’s knowledge. The data used for this research work contains 226 article’s records taken from the German version of Wikipedia. The dataset consists of 19 inputs and one output. The data was preprocessed to remove any similar attributes. The input variables are related to the editors, contributors, length of articles and the lifecycle of articles. In the end analysis of different methods implemented in this research is made to analyze the performance of each classification method used.

Η ραγδαία ανάπτυξη και διείσδυση των νέων τεχνολογιών πληροφορίας και επικοινωνίας έχει επιφέρει ριζικές αλλαγές σε όλους τους τομείς της ανθρώπινης δράσης (Castells, 1998). Ιδιαίτερο ενδιαφέρον παρουσιάζει η επιρροή των τεχνολογιών αυτών στον τομέα της εκπαίδευσης. Οι εξελίξεις στον χώρο της τεχνολογίας και επικοινωνίας καθώς και η διάδοση του Internet μετεξέλιξαν αναπόφευκτα την εκπαιδευτική διαδικασία, από το κλασσικό συγκεντρωτικό μοντέλο σε ένα πιο άμεσο και ευέλικτο: η «εξ’ Αποστάσεως Εκπαίδευση» (e-learning) είναι μια εναλλακτική μορφή εκπαίδευσης, που επιδιώκει να καλύψει τους περιορισμούς της παραδοσιακής εκπαίδευσης. Στην παρούσα μεταπτυχιακή διπλωματική εργασία σχεδιάστηκε και υλοποιήθηκε ένα ολοκληρωμένο σύστημα Δυναμικής Ανάλυσης και Πρόβλεψης της επίδοσης των εκπαιδευομένων, για ένα σύστημα εξ΄ αποστάσεως εκπαίδευσης. Η βασική ιδέα εμφορείται από την ανάγκη των ιδρυμάτων εξ΄ αποστάσεως εκπαίδευσης, για την κάλυψη των εκπαιδευτικών αναγκών και την παροχή υψηλής ποιότητας σπουδών. Η εξόρυξη γνώσης για την πρόβλεψη της επίδοσης των εκπαιδευομένων συμβάλλει καθοριστικά στην επίτευξη υψηλής ποιότητας σπουδών. Η ικανότητα και η δυνατότητα πρόβλεψης της απόδοσης των εκπαιδευομένων μπορεί να φανεί χρήσιμη με αρκετούς τρόπους για την διαμόρφωση ενός συστήματος, που θα μπορεί να αποτρέψει την αποτυχία καθώς και την παραίτηση των εκπαιδευομένων. Αξίζει να σημειωθεί ότι στα συστήματα εξ’ αποστάσεως εκπαίδευσης η συχνότητα «εγκατάλειψης» είναι αρκετά υψηλότερη από αυτή στα συμβατικά πανεπιστήμια. Για την πρόβλεψη της επίδοσης των εκπαιδευομένων, η απαιτούμενη πληροφορία βρίσκεται «κρυμμένη» στο εκπαιδευτικό σύνολο δεδομένων (δλδ. βαθμοί γραπτών εργασιών, βαθμοί τελικής εξέτασης, παρουσίες φοιτητών) και είναι εξαγώγιμη με τεχνικές εξόρυξης. Η χρήση μεθόδων εξόρυξης δεδομένων (data mining) στον τομέα της εκπαίδευσης παρουσιάζει αυξανόμενο ερευνητικό ενδιαφέρον. Ο νέος αυτός «αναπτυσσόμενος» τομέας έρευνας, που ονομάζεται «Εκπαιδευτική Εξόρυξη Δεδομένων», ασχολείται με την ανάπτυξη μεθόδων εξόρυξης «γνώσης» από τα εκπαιδευτικά σύνολα δεδομένων. Πράγμα που επιτυγχάνεται με τη χρήση τεχνικών όπως τα δέντρα απόφασης, τα Νευρωνικά Δίκτυα, Naïve Bayes, k-means, κλπ. Η παρούσα εργασία έχει σχεδιαστεί να προσφέρει ένα μοντέλο εξόρυξης δεδομένων χρησιμοποιώντας τη μέθοδο των δέντρων απόφασης, για το σύστημα τριτοβάθμιας εκπαίδευσης στο ανοιχτό πανεπιστήμιο. Η «γνώση» που προκύπτει από τα δεδομένα εξόρυξης θα χρησιμοποιηθεί με στόχο την διευκόλυνση και την ενίσχυση της μάθησης, καθώς επίσης και στη λήψη αποφάσεων. Στην παρούσα εργασία, εξάγουμε «γνώση» που σχετίζεται με τις επιδόσεις των μαθητών στην τελική εξέταση. Επίσης, γίνεται εντοπισμός των ατόμων που εγκαταλείπουν το μάθημα και των μαθητών που χρειάζονται ιδιαίτερη προσοχή και εντέλει δίνει τη δυνατότητα στους καθηγητές να παράσχουν την κατάλληλη παροχή συμβουλών. / The rapid development and intrusion of information technology and communications have caused radical changes in all sectors of human’s activity. (Castells, 1998). Of particular interest is the great technology’s influence on education. Due to the adoption of the new technologies, e-learning has been emerged and developed. As a result, distance learning has transformed and new possibilities have appeared. It is remarkable that distance learning became and considered as a scout of the new era in education and contributed to the quality of education: e-learning is trying to cover the limitations of conventional teaching environment. In the present thesis, an integrated system of dynamic analysis and prediction of the performance of students in distance education has been designed and implemented. The initial idea for designing this system came from the higher distance education institutes’ need to provide quality education to its students and to improve the quality of managerial decisions. One way to achieve highest level of quality in higher distance education e-learning system is by discovering knowledge from educational data to study the main attributes that may affect the students’ performance. The discovered knowledge can be used to offer a helpful and constructive recommendations to the academic planners in higher distance education institutes to enhance their decision making process, to improve students’ academic performance, trim down failure rate and dropout rate, to assist instructors, to improve teaching and many other benefits. Dropout rates in university level distance learning are definitely higher than those inconventional universities, thus limiting dropout is essential in university-level distance learning.

Η ραγδαία εξάπλωση του Internet έχει προκαλέσει σημαντικές αλλαγές σε πολλούς κλάδους της οικονομίας και της κοινωνίας παγκόσμια. Με τη ραγδαία ανάπτυξη των Τεχνολογιών της Πληροφορικής και της Τεχνολογίας, μια νέα μορφή εκπαίδευσης εμφανίστηκε, που δεν είναι άλλη από το e-learning (εκπαίδευση από απόσταση), που έφερε την επανάσταση στο εκπαιδευτικό γίγνεσθαι. Επιπρόσθετα ο Παγκόσμιος Ιστός σταδιακά μετεξελίσσεται στο Σημασιολογικό Παγκόσμιο Ιστό (Semantic Web) νέα μοντέλα και πρότυπα (XML, RDF, OWL) αναπτύσσονται για την προώθηση αυτής της διαδικασίας. Η έκφραση, μετάδοση και αναζήτηση πληροφοριών με χρήση αυτών των προτύπων ανοίγει νέους ορίζοντες στη χρήση του Διαδικτύου. Οι οντολογίες κερδίζουν ολοένα έδαφος για την αναπαράσταση γνώσης. Σε μια μεγάλη οντολογία που περιέχει χρήσιμα δεδομένα για ένα σύστημα εξ’ αποστάσεως εκπαίδευσης, αξίζει κάποιος να ερευνήσει την «κρυμμένη γνώση», δηλαδή να ανακαλύψει πιθανές συσχετίσεις ή συνειρμούς, να βρει πρότυπα ή μορφές που επαναλαμβάνονται ή ακραία φαινόμενα. Η παρούσα διπλωματική εργασία αποτελεί μια επίδειξη τεχνολογίας για την έγκυρη και έγκαιρη πρόβλεψη της απόδοσης των φοιτητών σε ένα σύστημα εξ’ αποστάσεως εκπαίδευσης. Η βασική ιδέα προκύπτει από την ανάγκη να σχεδιαστεί μία οντολογία η οποία θα μπορεί να αποθηκεύσει τη γνώση σχετικά με τις ικανότητες φοιτητών (user profile) σε σχέση με ένα συγκεκριμένο εκπαιδευτικό αντικείμενο (ΠΛΗ23 – Τηλεματική, Διαδίκτυο του Ελληνικού Ανοικτού Πανεπιστημίου (ΕΑΠ) )η οποία έχει πολύ συγκεκριμένη ύλη και 4 υποχρεωτικές γραπτές εργασίες ανά έτος). Στη συνέχεια παρουσιάζονται τα αποτελέσματα μελέτης της ανάλυσης των δεδομένων των φοιτητών με τεχνικές εξόρυξης γνώσης. Η εύρεση των κανόνων πραγματοποιήθηκε μέσω του εργαλείου Weka. Το αποτέλεσμα που προέκυψε είναι μία βάση γνώσης βάσει της οποίας γίνεται έγκαιρα και έγκυρα η πρόβλεψη της συμπεριφοράς του φοιτητή, δηλαδή αν θα καταφέρει να ολοκληρώσει επιτυχώς ή μη τη Θεματική Ενότητα που έχει αναλάβει στο ΕΑΠ, ώστε ο διδάσκων να μπορεί από πολύ νωρίς να υποστηρίξει το φοιτητή με επιπλέον υλικό αν απαιτείται. / The rapid spread of Internet has caused significant changes in many sectors of the economy and society worldwide. From those changes could not be left out of education. With the rapid development of information technologies and technology, a new form of education appears, e-learning (distance education), which revolutionized the educational process. Furthermore, while the World Wide Web gradually transforms into Semantic Web, new standards and models (XML, RDF, OWL) are evolving in order to launch this inquiry. The storage, presentation, transmission and search of information according to those standards open up new horizons in the utilization of the Web. Ontologies are increasingly get used for knowledge representation. A large ontology contains useful data for a system of distance education, deserves someone to investigate the "hidden knowledge", i.e. to discover possible associations or to find patterns or forms that are repeated or extreme events. This thesis is a demonstration of technology for accurate and timely prediction of the performance of students in a system of distance education. The basic idea was to design an ontology that can store knowledge about the students’ skills (user profile) in relation to a specific educational purpose (PLI23 - Telematics, Internet of the Hellenic Open University, which has a very specific matter and 4 mandatory projects per year). Then we present the results of a study analyzing student data mining techniques (data mining-classification). The discovery rules took place via the tool Weka. The result is a knowledge base which is the appropriate tool (Interface teacher) may provide that a student needs on a particular topic (in addition to material help from the teacher), etc.

The interest in social robots has grown dramatically in the last decade. Several studies have investigated the potential markets for such robots and how to enhance their human-like abilities. Both of these subjects have been investigated in this thesis using the company Furhat Robotics, and their robot Furhat, as a case study. This paper explores how machine learning could be used to classify dialogue acts in human-robot conversations, which could help Furhat interact in a more human-like way. Dialogue acts are acts of natural speech, such as questions or statements. Several variables and their impact on the classification of dialogue acts were tested. The results showed that a combination of some of these variables could classify 73 % of all the dialogue acts correctly. Furthermore, this paper analyzes the market for social robots which are used for education, where human-like abilities are preferable. A literature study and an interview were conducted. The market was then analyzed using a SWOT-matrix and Porter’s Five Forces. Although the study showed that the mentioned market could be a suitable target for Furhat Robotics, there are several threats and obstacles that should be taken into account before entering the market. / Intresset för sociala robotar har ökat drastiskt under det senaste årtiondet. Ett flertal studier har undersökt hur man kan förbättra robotars mänskliga färdigheter. Vidare har studier undersökt potentiella marknader för sådana robotar. Båda dessa aspekter har studerats i denna rapport med företaget Furhat Robotics, och deras robot Furhat, som en fallstudie. Mer specifikt undersöker denna rapport hur maskininlärning kan användas för att klassificera talhandlingar i människa-robot- konversationer, vilket skulle kunna hjälpa Furhat att interagera på ett mer mänskligt sätt. Talhandlingar är indelningar av naturligt språk i olika handlingar, såsom frågor och påståenden. Flertalet variabler och deras inverkan på klassificeringen av talhandlingar testades i studien. Resultatet visade att en kombination av några av dessa variabler kunde klassificera 73 % av alla talhandlingar korrekt. Vidare analyserar denna rapport marknaden för sociala robotar inom utbildning, där mänskliga färdigheter är att föredra. En litteraturstudie och en intervju gjordes. Marknaden analyserades sedan med hjälp av en SWOT-matris och Porters femkraftsmodell. Fastän studien visade att den ovannämnda marknaden skulle kunna vara lämplig för Furhat Robotics finns ett flertal hot och hinder som företaget måste ta hänsyn till innan de tar sig in på marknaden.

Submitted by admin tede (tede@pucgoias.edu.br) on 2016-11-10T11:48:54Z No. of bitstreams: 1 Willard Silva Ribeiro.pdf: 2550243 bytes, checksum: 6a14ee1c5b586e0f46c616a14e6fe721 (MD5) / Made available in DSpace on 2016-11-10T11:48:54Z (GMT). No. of bitstreams: 1 Willard Silva Ribeiro.pdf: 2550243 bytes, checksum: 6a14ee1c5b586e0f46c616a14e6fe721 (MD5) Previous issue date: 2016-09-12 / This work deals with the optimization of processes that result in improved financial and sales of telecommunications services in the State of Goiás through the analysis of undesirable effects (UDEs) systematized by the Theory of Constraints (TOC) combined with the use of data mining in fixed-line and mobile channels. In mobile phone retail channels, mass, own and franchise stores are analyzed. In the wireline phones segment we analyze authorized agent and door to door. The Analysis of Unwanted effects (UE) is systematized in cause and effect and uses the theory of constraints through the Current Reality Tree tool (CRT) to understand the root problem, which would be the unwanted effect feeding and causing other unwanted effects in the set of processes of each sales channel. The evaporation cloud diagram (ECD), which was used to break the assumptions that kept the active restriction and enabled the creation of the injection that would transform UDES in desired effects(DE). The negative branch reservation (NBR) to reduce the risk of application Injection avoiding the creation of new UDES. And the future reality tree (FRT) that shows the improved model of the future system. For the classification of business segments is used data mining, the J48 algorithm through the WEKA software. The results show that the data mining and the theory of the constraints provides a better understanding of the processes, a better understanding of the consumers behavior in each sales segment and improving sales. / Esta pesquisa diz respeito a melhoria de processos que resultam em aumento no volume de vendas de serviços de telecomunicações em duas concessionárias de telecomunicações no Estado de Goiás por meio da análise de efeitos indesejados (Undesirable Effects, UDEs) sistematizados através de Teoria das Restrições (Theory of Constraints, TOC) aliado ao uso de mineração de dados (Data Mining, DM) nos canais de vendas (Sales Channel, SC) de telefonia fixa e telefonia móvel entre os anos de 2013 a 2015. Na telefonia móvel são analisados os canais de varejo, canais de vendas em massa, lojas próprias e franquias. Na telefonia fixa são analisados o canal agente autorizado e canal porta a porta. A análise dos UDEs é sistematizada em causa e efeito e utiliza a TOC através da ferramenta Árvore de Realidade Atual (Current Reality Tree, CRT) para compreensão do problema raiz que seria o UDEs que alimenta e causa outros UDEs no conjunto de processos de cada SC. A causa raiz é submetida ao Diagrama de Evaporação de Nuvem (Evaporiting Cloud Diagram, ECD), que foi usado para quebra dos pressupostos que mantinham a restrição ativa e possibilitou a criação da injeção que transformaria UDEs em Efeitos Desejados (Desirable Effects, DEs). A injeção, que é uma sentença que representa uma ação desruptiva do que causava a restrição, foi testada por meio da ferramenta Reserva de Ramo Negativo (Negative Branch Reservation, NBR) para reduzir os riscos da aplicação da Injeção evitando a criação de novos UDES. UDEs foram transformados em DES por meio da injeção validada e sistematizados em causa e efeito com uso da Árvore de Realidade Futura (Future Reality Tree, FRT) que apresenta o modelo melhorado futuro do sistema. Para a classificação dos segmentos comerciais é usado DM com o algoritmo de classificação J48 por meio do software Waikato Environment for Knowledge Analysis (WEKA). Os resultados mostram que a mineração de dados e a teoria das restrições oferecem uma melhor compreensão dos processos de vendas, melhor compreensão do comportamento de consumo em cada segmento comercial e por consequência a melhoria das vendas.

Thesis (M.S.)--Joint Program in Physical Oceanography (Massachusetts Institute of Technology, Dept. of Earth, Atmospheric, and Planetary Sciences; and the Woods Hole Oceanographic Institution), 1999. / Includes bibliographcial references (p. 117-118). / This thesis describes an attempt to build a box model of the circulation in the eastern North Atlantic and to use it with transient tracer data to infere rates of ventilation in the subtropcial thermocline. The starting point is the analysis of former studies on box models, transient tracer data and the combination of both. The advantages and pitfalls associated with the tracer age approach on the one hand and the inventory approach on the other hand are weighed and the choice set upon the inventory approach is justified. Next the data used is presented and processed, and the results are compared with the known circulation patterns of the basin. The estimates computed fall in the expected and acceptable range. The uncertainties, in particular on the boundary conditions, are acknowledged to be a crucial factor on the following analysis yet only rough estimates can be produced. In particular, the geostrophic velocities at the boundaries can only be determined lest an unknown constant. No internally consistent model can be found that satisfies the linear conservation balances, geostrophy and steadiness assumptions and the boundary conditions imposed. The circulation generated only satisfies mass balance and the boundary conditions to a certain extent. Experience suggests that there are incompatibilities among the various constraints. Two different numerical methods fail to find an acceptable solution. Using the default circulation obtained, the forward problem is formulated and investigated. The resulting tracer distribution and time history is incompatible with the observed field. As a consequence, an attempt is made at the inverse problem in the hope that relaxation of the boundary conditions will provide some insight into the general failure of the model. As there appears to be no feasible solution though, the circulation is further inspected and it is concluded that given its flaws, no boundary condition will be able to generate a tracer field even in partial agreement with the observations. It is finally concluded that transient tracers can be used to dismiss grossly wrong circulation models. / by Gwenaëlle C. Jeunhomme. / M.S.

Despite widespread recognition by science educators, researchers and K-12 frameworks that scientific inquiry should be an essential part of science education, typical classrooms and assessments still emphasize rote vocabulary, facts, and formulas. One of several reasons for this is that the rigorous assessment of complex inquiry skills is still in its infancy. Though progress has been made, there are still many challenges that hinder inquiry from being assessed in a meaningful, scalable, reliable and timely manner. To address some of these challenges and to realize the possibility of formative assessment of inquiry, we describe a novel approach for evaluating, tracking, and scaffolding inquiry process skills. These skills are demonstrated as students experiment with computer-based simulations. In this work, we focus on two skills related to data collection, designing controlled experiments and testing stated hypotheses. Central to this approach is the use and extension of techniques developed in the Intelligent Tutoring Systems and Educational Data Mining communities to handle the variety of ways in which students can demonstrate skills. To evaluate students' skills, we iteratively developed data-mined models (detectors) that can discern when students test their articulated hypotheses and design controlled experiments. To aggregate and track students' developing latent skill across activities, we use and extend the Bayesian Knowledge-Tracing framework (Corbett & Anderson, 1995). As part of this work, we directly address the scalability and reliability of these models' predictions because we tested how well they predict for student data not used to build them. When doing so, we found that these models demonstrate the potential to scale because they can correctly evaluate and track students' inquiry skills. The ability to evaluate students' inquiry also enables the system to provide automated, individualized feedback to students as they experiment. As part of this work, we also describe an approach to provide such scaffolding to students. We also tested the efficacy of these scaffolds by conducting a study to determine how scaffolding impacts acquisition and transfer of skill across science topics. When doing so, we found that students who received scaffolding versus students who did not were better able to acquire skills in the topic in which they practiced, and also transfer skills to a second topic when was scaffolding removed. Our overall findings suggest that computer-based simulations augmented with real-time feedback can be used to reliably measure the inquiry skills of interest and can help students learn how to demonstrate these skills. As such, our assessment approach and system as a whole shows promise as a way to formatively assess students' inquiry.

Despite widespread recognition by science educators, researchers and K-12 frameworks that scientific inquiry should be an essential part of science education, typical classrooms and assessments still emphasize rote vocabulary, facts, and formulas. One of several reasons for this is that the rigorous assessment of complex inquiry skills is still in its infancy. Though progress has been made, there are still many challenges that hinder inquiry from being assessed in a meaningful, scalable, reliable and timely manner. To address some of these challenges and to realize the possibility of formative assessment of inquiry, we describe a novel approach for evaluating, tracking, and scaffolding inquiry process skills. These skills are demonstrated as students experiment with computer-based simulations. In this work, we focus on two skills related to data collection, designing controlled experiments and testing stated hypotheses. Central to this approach is the use and extension of techniques developed in the Intelligent Tutoring Systems and Educational Data Mining communities to handle the variety of ways in which students can demonstrate skills. To evaluate students' skills, we iteratively developed data-mined models (detectors) that can discern when students test their articulated hypotheses and design controlled experiments. To aggregate and track students' developing latent skill across activities, we use and extend the Bayesian Knowledge-Tracing framework (Corbett & Anderson, 1995). As part of this work, we directly address the scalability and reliability of these models' predictions because we tested how well they predict for student data not used to build them. When doing so, we found that these models demonstrate the potential to scale because they can correctly evaluate and track students' inquiry skills. The ability to evaluate students' inquiry also enables the system to provide automated, individualized feedback to students as they experiment. As part of this work, we also describe an approach to provide such scaffolding to students. We also tested the efficacy of these scaffolds by conducting a study to determine how scaffolding impacts acquisition and transfer of skill across science topics. When doing so, we found that students who received scaffolding versus students who did not were better able to acquire skills in the topic in which they practiced, and also transfer skills to a second topic when was scaffolding removed. Our overall findings suggest that computer-based simulations augmented with real-time feedback can be used to reliably measure the inquiry skills of interest and can help students learn how to demonstrate these skills. As such, our assessment approach and system as a whole shows promise as a way to formatively assess students' inquiry.