izpis_h1_title_alt

Model adaptivnega sistema za priporočanje učnih objektov v konstruktivističnem učnem okolju
Ocepek, Uroš (Author), Bosnić, Zoran (Mentor) More about this mentor... This link opens in a new window, Rugelj, Jože (Mentor) More about this mentor... This link opens in a new window

URLURL - Presentation file, Visit http://eprints.fri.uni-lj.si/3207/ This link opens in a new window

Abstract
Smernice sodobnega učenja kažejo na razvoj prilagodljivih spletnih učnih okolij, ki prilagajajo učno vsebino učenčevim lastnostim in potrebam. Priporočilni sistemi lahko predstavljajo takšne sisteme, saj temeljijo na priporočanju ustreznih učnih objektov glede na uporabnikove lastnosti, kot so: učne karakteristike, predhodna ocenjevanja objektov in demografske lastnosti. Če za uporabnika nimamo začetnih ocen, pravimo, da je uporabnik v stanju hladnega zagona. Tema disertacije je oblikovanje koncepta priporočilnega sistema, ki vsebuje nov pristop k reševanju problema hladnega zagona. Na splošno se pristopi za reševanje tega problema delijo na dve skupini: pristope, ki temeljijo na začetnem poizvedovanju o uporabnikovih lastnostih, in pristope, ki uporabljajo specifične algoritme za uporabnike, ki so v stanju hladnega zagona. Najprej smo oblikovali pristop, ki sodi v prvo skupino pristopov – začetno poizvedovanje. Z namenom, da bi učencem priporočili čim bolj ustrezna multimedijska gradiva, smo povezali različne učne stile s posameznimi multimedijskimi tipi. Povezali smo Kolbov model učnih stilov, Rancourtov model, hemisferični model in model učnih stilov VAK (vizualni, slušni in kinestetični stil). Zgradili in analizirali smo odločitveni model za priporočanje multimedijskega gradiva. Zgrajeni model je pokazal, da je hemisferični model učnih stilov najpomembnejši odločitveni atribut, Kolbov model in model VAK pa sta druga najpomembnejša odločitvena atributa. Raziskava je pokazala, da študentje za učenje najpogosteje uporabljajo besedila, ki so dobro strukturirana in imajo barvno diskriminacijo. Poleg besedil študentje pogosto uporabljajo animacije in videoposnetke. Z zgrajenim odločitvenim modelom smo tudi potrdili, da je za ustreznejšo izbiro in priporočanje različnih tipov multimedijskih gradiv smiselno kombinirati različne modele učnih stilov. Po drugi strani pa je model statičen, kar pomeni, da v vsaki situaciji priporoča enak multimedijski tip. To ni ustrezno, ker ni nujno, da je neki multimedijski tip ustrezen za vse učne teme. Drugi del raziskovanja je bil namenjen oblikovanju in razvoju inovativnega pristopa, ki izboljša priporočanje uporabnikom v hladnem zagonu, in sicer tako, da uporabnikom vstavi samo nekaj ocen za izbrane objekte in nato nad celotno matriko izvede matrično faktorizacijo. Naš pristop je sestavljen iz treh korakov: 1) iskanje podobnih uporabnikov; 2) izbor ciljnih atributov; 3) agregacija pridobljene vrednosti v izbrane ciljne atribute. Reševanje hladnega zagona smo razdelili na reševanje problema absolutnega in delnega hladnega zagona. Glede na rezultate eksperimenta (reševanje problema absolutnega hladnega zagona) so rezultati pokazali, da predlagane metode v povprečju izboljšajo nenegativno matrično faktorizacijo s stohastičnim gradientnim spustom (NG)}. Pri polnenegativni matrični faktorizaciji z manjkajočimi vrednostmi (SN) pa kombinaciji FR-ME (v atribut, ki smo ga izbrali glede na frekvenco najpogostejše vrednosti, vstavimo povprečno vrednost vseh ocen v atributu) in SD-MF (v atribut, ki smo ga izbrali glede na vrednost standardnega odklona, vstavimo najpogostejšo vrednost atributa) bistveno izboljšata priporočanje. Pri preostalih dveh matričnih faktorizacijah (matrična faktorizacija z zlivanjem podatkovnih virov (DF) in nenegativna matrična faktorizacija z izmenjujočimi se najmanjšimi kvadrati (NS)) predlagane kombinacije ne izboljšajo priporočanja. V nadaljevanju smo naš pristop testirali za reševanje delnega hladnega zagona. Statistična analiza eksperimentalnega dela (testiranja našega pristopa) na umetni množici je pokazala, da vsak posamezen parameter statistično značilno izboljša priporočanje metod matričnih faktorizacij. Najboljše delovanje so pokazale metode, pri katerih upoštevamo 25 % najbližjih sosedov ($25$-*-*-*), uporabljamo frekvenco (*-FR-*-*) ali RReliefF (*-RR-*-*) za izbiro atributov in vstavljamo vrednosti s povprečenjem (*-*-ME-*) ali z uporabo regresijskega drevesa (*-*-RT-*). V nadaljevanju smo izbrali varianti metod, za kateri smo pričakovali, da bosta dajali dobre rezultate: 25-FR-ME-* in 25-RR-RT-*. Izbranima variantama metod smo dodali še kombinacijo, pri kateri upoštevamo 50 % najbližjih sosedov, za izbor atributov uporabimo RReliefF in vstavljamo vrednosti z uporabo linearne regresije ($50$-RR-LR-*) – kombinacija je bila glede na povprečni rang najboljša. Vse tri izbrane variante metod smo testirali na dveh realnih množicah: Jester in PEFbase. Evalvacija je pokazala, da vstavljanje manjkajočih vrednosti s predlagano metodo 25-FR-ME-* in uporabo matrične faktorizacije NG značilno izboljša priporočanje objektov v primerjavi z golimi metodami matrične faktorizacije (DF, NG, NS in SN), in sicer za uporabnike v stanju delnega hladnega zagona. Preverili smo tudi vpliv vstavljanja manjkajočih ocen na druge uporabnike. Ugotovili smo, da naše spremembe ne povzročajo značilnih sprememb pri priporočanju drugim uporabnikom, ki so zunaj stanja delnega hladnega zagona.

Language:Unknown
Keywords:vstavljanje manjkajočih vrednosti, problem novega uporabnika, matrična faktorizacija, priporočilni sistemi
Work type:Doctoral dissertation (mb31)
Organization:FRI - Faculty of computer and information science
Year:2015
Views:1088
Downloads:179
Metadata:XML RDF-CHPDL DC-XML DC-RDF
 
Average score:(0 votes)
Your score:Voting is allowed only to logged in users.
:
Share:AddThis
AddThis uses cookies that require your consent. Edit consent...

Secondary language

Language:Unknown
Title:A model of an adaptive system for recommending learning objects in a constructivist learning environment
Abstract:
Computer-based multimedia learning environments support the idea that people learn better and more deeply when appropriate pictures (i.e., animations, video, static graphics) are added to text or narration. There are many adaptive learning systems that adapt learning materials to student properties, preferences, and activities. Adaptive learning environments mostly support only traditional concepts of learning. There is a need to design and develop an e-learning system that embodies principles of constructivist learning approach. The solution is in recommenders systems, which suggest items of interest to users based on their preferences (i.e. previous ratings). If there are no ratings for a certain user or item/object, there is a situation called a cold start problem, which leads to unreliable recommendations. Researchers mostly avoid tackling the absolute cold start in recommender systems. The topic of presented dissertation is designing a recommender system with a novel approach to avoid cold start problem. Approaches for solving the new user cold start problem can be divided into two main groups: the first group performs additional inquiries to gather more information about the users; and the second group uses dedicated algorithms for users in the cold start state. The first group of approaches aims at performing additional inquiries about the user. According to this approach, we relate combinations of different learning styles (taking into account four different learning styles models) to preferred multimedia types. We explore a decision model aimed at proposing learning material of an appropriate multimedia type. The study includes 272 student participants. The resulting decision model shows that students prefer well-structured learning texts with colour discrimination, and that the hemispheric learning style model is the most important criterion in deciding student preferences for different multimedia learning materials. To provide a more accurate and reliable model for recommending different multimedia types more learning style models must be combined. Kolb’s classification and the VAK classification allow us to learn if students prefer an active role in the learning process, and what multimedia type they prefer. The results also shows that there is an obvious need to combine learning styles model in order to get a wider view of the student’s characteristics: an approach to problem solving problems, cognitive modes, way of thinking, and a dominant mode of perceiving information. On another hand, model recommends same multimedia material regardless of the learning topic. In the second part of our research, we have designed and developed a novel approach for alleviating the cold start problem by imputing missing values into the input matrix, thereby improving recommendation performance. Our approach has three steps: 1) finding similar users to given user in cold start state; 2) selecting relevant attributes for the imputation process; 3) aggregate ratings to input matrix for a user in the cold start state. We separate our approach for solving cold start problem into solving absolute cold start problem and solving partial cold start problem. According to the results of our experiments (solving absolute cold start problem), the results indicate that all our proposed methods improve recommending for non-negative matrix factorization with stochastic gradient descent (NG). For semi-non-negative matrix factorization with missing data (SN), combinations FR-ME (imputing attribute's mean value into the attributes that have the highest frequency of the most frequent values) and SD-MF (imputing attribute’s most frequent value into attributes that have the lowest standard deviation) improve recommendations for users in the absolute cold start state. For non-negative matrix factorization with alternating least squares (NS) and matrix factorization by data fusion (DF), none of variations of proposed parameters (methods) improves recommending in absolute cold start state. In the next stage of our research, we evaluated our approach for solving partial cold start problem. Statistical analysis of experimental evaluation of our approach on the artificial domain showed that each parameter significantly improved recommending of matrix factorization methods. The methods that yield improvements in recommendation accuracy compared with the raw matrix factorization are methods that consider 25 % of similar users ($25$-*-*-*), select an attribute according to the frequency (*-FR-*-*) or RReliefF (*-RR-*-*), and impute a value aggregated by mean value (ME) or predicted by using regression trees (RT). For further investigation we chose two method combinations (25-FR-ME-* and 25-RR-RT-*), which were expected to work well, and compared them with other strategies on real domains. Among all approaches evaluated on the artificial domain, we chose the best performing method with the highest average rank – a method that considers 50 % of similar users, selects an attribute for imputation according to the RReliefF, and imputes a value predicted by linear regression (50-RR-LR-*). All three combinations of the selected methods were evaluated on two real domains: Jester in PEFbase. An evaluation showed that method 25-FR-ME-* combined with matrix factorization NG performed statistically better than the raw matrix factorization algorithms (DF, NG, NS in SN) on real domains for users in the partial cold-start state. The results demonstrated the advantage of using imputation approaches in terms of better recommendation accuracy. At the same time, the results have shown that imputing of missing values has no negative impact for recommending to the users, which are not in the cold start state.

Keywords:imputing missing values, new user problem, matrix factorization, recommender systems

Similar documents

Similar works from RUL:
Similar works from other Slovenian collections:

Comments

Leave comment

You have to log in to leave a comment.

Comments (0)
0 - 0 / 0
 
There are no comments!

Back