Müllner Peter , Lex Elisabeth, Schedl Markus, Kowald Dominik
2024
Collaborative filtering-based recommender systems leverage vast amounts of behavioral user data, which poses severe privacy risks. Thus, often random noise is added to the data to ensure Differential Privacy (DP). However, to date it is not well understood in which ways this impacts personalized recommendations. In this work, we study how DP affects recommendation accuracy and popularity bias when applied to the training data of state-of-the-art recommendation models.Our findings are three-fold: First, we observe that nearly all users' recommendations change when DP is applied. Second, recommendation accuracy drops substantially while recommended item popularity experiences a sharp increase, suggesting that popularity bias worsens. Finally, we find that DP exacerbates popularity bias more severely for users who prefer unpopular items than for users who prefer popular items.
Marta Moscati, Christian Wallman, Markus Reiter-Haas, Kowald Dominik, Elisabeth Lex, Markus Schedl
2023
Integrating the ACT-R Framework with Collaborative Filtering for Explainable Sequential Music Recommendati
Kowald Dominik, Gregor Mayr, Markus Schedl, Elisabeth Lex
2023
A Study on Accuracy, Miscalibration, and Popularity Bias in Recommendation
Lacic Emanuel, Duricic Tomislav, Fadljevic Leon, Theiler Dieter, Kowald Dominik
2023
Uptrendz: API-Centric Real-Time Recommendations in Multi-Domain Settings
Müllner Peter , Schmerda Stefan, Theiler Dieter, Lindstaedt Stefanie , Kowald Dominik
2022
Data and algorithm sharing is an imperative part of data- and AI-driven economies. The efficient sharing of data and algorithms relies on the active interplay between users, data providers, and algorithm providers. Although recommender systems are known to effectively interconnect users and items in e-commerce settings, there is a lack of research on the applicability of recommender systems for data and algorithm sharing. To fill this gap, we identify six recommendation scenarios for supporting data and algorithm sharing, where four of these scenarios substantially differ from the traditional recommendation scenarios in e-commerce applications. We evaluate these recommendation scenarios using a novel dataset based on interaction data of the OpenML data and algorithm sharing platform, which we also provide for the scientific community. Specifically, we investigate three types of recommendation approaches, namely popularity-, collaboration-, and content-based recommendations. We find that collaboration-based recommendations provide the most accurate recommendations in all scenarios. Plus, the recommendation accuracy strongly depends on the specific scenario, e.g., algorithm recommendations for users are a more difficult problem than algorithm recommendations for datasets. Finally, the content-based approach generates the least popularity-biased recommendations that cover the most datasets and algorithms.
Lacic Emanuel, Kowald Dominik
2022
In this industry talk at ECIR'2022, we illustrate how to build a modern recommender system that can serve recommendations in real-time for a diverse set of application domains. Specifically, we present our system architecture that utilizes popular recommendation algorithms from the literature such as Collaborative Filtering, Content-based Filtering as well as various neural embedding approaches (e.g., Doc2Vec, Autoencoders, etc.). We showcase the applicability of our system architecture using two real-world use-cases, namely providing recommendations for the domains of (i) job marketplaces, and (ii) entrepreneurial start-up founding. We strongly believe that our experiences from both research- and industry-oriented settings should be of interest for practitioners in the field of real-time multi-domain recommender systems.
Lacic Emanuel, Fadljevic Leon, Weissenböck Franz, Lindstaedt Stefanie , Kowald Dominik
2022
Personalized news recommender systems support readers in finding the right and relevant articles in online news platforms. In this paper, we discuss the introduction of personalized, content-based news recommendations on DiePresse, a popular Austrian online news platform, focusing on two specific aspects: (i) user interface type, and (ii) popularity bias mitigation. Therefore, we conducted a two-weeks online study that started in October 2020, in which we analyzed the impact of recommendations on two user groups, i.e., anonymous and subscribed users, and three user interface types, i.e., on a desktop, mobile and tablet device. With respect to user interface types, we find that the probability of a recommendation to be seen is the highest for desktop devices, while the probability of interacting with recommendations is the highest for mobile devices. With respect to popularity bias mitigation, we find that personalized, content-based news recommendations can lead to a more balanced distribution of news articles' readership popularity in the case of anonymous users. Apart from that, we find that significant events (e.g., the COVID-19 lockdown announcement in Austria and the Vienna terror attack) influence the general consumption behavior of popular articles for both, anonymous and subscribed users
Kowald Dominik, Lacic Emanuel
2022
Multimedia recommender systems suggest media items, e.g., songs, (digital) books and movies, to users by utilizing concepts of traditional recommender systems such as collaborative filtering. In this paper, we investigate a potential issue of such collaborative-filtering based multimedia recommender systems, namely popularity bias that leads to the underrepresentation of unpopular items in the recommendation lists. Therefore, we study four multimedia datasets, i.e., LastFm, MovieLens, BookCrossing and MyAnimeList, that we each split into three user groups differing in their inclination to popularity, i.e., LowPop, MedPop and HighPop. Using these user groups, we evaluate four collaborative filtering-based algorithms with respect to popularity bias on the item and the user level. Our findings are three-fold: firstly, we show that users with little interest into popular items tend to have large user profiles and thus, are important data sources for multimedia recommender systems. Secondly, we find that popular items are recommended more frequently than unpopular ones. Thirdly, we find that users with little interest into popular items receive significantly worse recommendations than users with medium or high interest into popularity.
Duricic Tomislav, Kowald Dominik, Schedl Markus, Lex Elisabeth
2021
Homophily describes the phenomenon that similarity breeds connection, i.e., individuals tend to form ties with other people who are similar to themselves in some aspect(s). The similarity in music taste can undoubtedly influence who we make friends with and shape our social circles. In this paper, we study homophily in an online music platform Last.fm regarding user preferences towards listening to mainstream (M), novel (N), or diverse (D) content. Furthermore, we draw comparisons with homophily based on listening profiles derived from artists users have listened to in the past, i.e., artist profiles. Finally, we explore the utility of users' artist profiles as well as features describing M, N, and D for the task of link prediction. Our study reveals that: (i) users with a friendship connection share similar music taste based on their artist profiles; (ii) on average, a measure of how diverse is the music two users listen to is a stronger predictor of friendship than measures of their preferences towards mainstream or novel content, i.e., homophily is stronger for D than for M and N; (iii) some user groups such as high-novelty-seekers (explorers) exhibit strong homophily, but lower than average artist profile similarity; (iv) using M, N and D achieves comparable results on link prediction accuracy compared with using artist profiles, but the combination of features yields the best accuracy results, and (v) using combined features does not add value if graph-based features such as common neighbors are available, making M, N, and D features primarily useful in a cold-start user recommendation setting for users with few friendship connections. The insights from this study …
Müllner Peter , Kowald Dominik, Lex Elisabeth
2021
In this paper, we explore the reproducibility of MetaMF, a meta matrix factorization framework introduced by Lin et al. MetaMF employs meta learning for federated rating prediction to preserve users' privacy. We reproduce the experiments of Lin et al. on five datasets, i.e., Douban, Hetrec-MovieLens, MovieLens 1M, Ciao, and Jester. Also, we study the impact of meta learning on the accuracy of MetaMF's recommendations. Furthermore, in our work, we acknowledge that users may have different tolerances for revealing information about themselves. Hence, in a second strand of experiments, we investigate the robustness of MetaMF against strict privacy constraints. Our study illustrates that we can reproduce most of Lin et al.'s results. Plus, we provide strong evidence that meta learning is essential for MetaMF's robustness against strict privacy constraints.
Duricic Tomislav, Hussain Hussain, Lacic Emanuel, Kowald Dominik, Lex Elisabeth, Helic Denis
2020
In this work, we study the utility of graph embeddings to generate latent user representations for trust-based collaborative filtering. In a cold-start setting, on three publicly available datasets, we evaluate approaches from four method families:(i) factorization-based,(ii) random walk-based,(iii) deep learning-based, and (iv) the Large-scale Information Network Embedding (LINE) approach. We find that across the four families, random-walk-based approaches consistently achieve the best accuracy. Besides, they result in highly novel and diverse recommendations. Furthermore, our results show that the use of graph embeddings in trust-based collaborative filtering significantly improves user coverage
Kowald Dominik, Markus Schedl, Lex Elisabeth
2020
Research has shown that recommender systems are typicallybiased towards popular items, which leads to less popular items beingunderrepresented in recommendations. The recent work of Abdollahpouriet al. in the context of movie recommendations has shown that this pop-ularity bias leads to unfair treatment of both long-tail items as well asusers with little interest in popular items. In this paper, we reproducethe analyses of Abdollahpouri et al. in the context of music recommen-dation. Specifically, we investigate three user groups from the Last.fmmusic platform that are categorized based on how much their listen-ing preferences deviate from the most popular music among all Last.fmusers in the dataset: (i) low-mainstream users, (ii) medium-mainstreamusers, and (iii) high-mainstream users. In line with Abdollahpouri et al.,we find that state-of-the-art recommendation algorithms favor popularitems also in the music domain. However, their proposed Group Aver-age Popularity metric yields different results for Last.fm than for themovie domain, presumably due to the larger number of available items(i.e., music artists) in the Last.fm dataset we use. Finally, we comparethe accuracy results of the recommendation algorithms for the three usergroups and find that the low-mainstreaminess group significantly receivesthe worst recommendations.
Fadljevic Leon, Maitz Katharina, Kowald Dominik, Pammer-Schindler Viktoria, Gasteiger-Klicpera Barbara
2020
This paper describes the analysis of temporal behavior of 11--15 year old students in a heavily instructionally designed adaptive e-learning environment. The e-learning system is designed to support student's acquisition of health literacy. The system adapts text difficulty depending on students' reading competence, grouping students into four competence levels. Content for the four levels of reading competence was created by clinical psychologists, pedagogues and medicine students. The e-learning system consists of an initial reading competence assessment, texts about health issues, and learning tasks related to these texts. The research question we investigate in this work is whether temporal behavior is a differentiator between students despite the system's adaptation to students' reading competence, and despite students having comparatively little freedom of action within the system. Further, we also investigated the correlation of temporal behaviour with performance. Unsupervised clustering clearly separates students into slow and fast students with respect to the time they take to complete tasks. Furthermore, topic completion time is linearly correlated with performance in the tasks. This means that we interpret working slowly in this case as diligence, which leads to more correct answers, even though the level of text difficulty matches student's reading competence. This result also points to the design opportunity to integrate advice on overarching learning strategies, such as working diligently instead of rushing through, into the student's overall learning activity. This can be done either by teachers, or via additional adaptive learning guidance within the system.
Duricic Tomislav, Lacic Emanuel, Kowald Dominik, Lex Elisabeth
2019
User-based Collaborative Filtering (CF) is one of the most popular approaches to create recommender systems. CF, however, suffers from data sparsity and the cold-start problem since users often rate only a small fraction of available items. One solution is to incorporate additional information into the recommendation process such as explicit trust scores that are assigned by users to others or implicit trust relationships that result from social connections between users. Such relationships typically form a very sparse trust network, which can be utilized to generate recommendations for users based on people they trust. In our work, we explore the use of regular equivalence applied to a trust network to generate a similarity matrix that is used for selecting k-nearest neighbors used for item recommendation. Two vertices in a network are regularly equivalent if their neighbors are themselves equivalent and by using the iterative approach of calculating regular equivalence, we can study the impact of strong and weak ties on item recommendation. We evaluate our approach on cold start users on a dataset crawled from Epinions and find that by using weak ties in addition to strong ties, we can improve the performance of a trust-based recommender in terms of recommendation accuracy.
Kowald Dominik, Lex Elisabeth, Schdel Markus
2019
Kopeinik Simone, Lex Elisabeth, Kowald Dominik, Albert Dietrich, Seitlinger Paul
2019
When people engage in Social Networking Sites, they influence one another through their contributions. Prior research suggests that the interplay between individual differences and environmental variables, such as a person’s openness to conflicting information, can give rise to either public spheres or echo chambers. In this work, we aim to unravel critical processes of this interplay in the context of learning. In particular, we observe high school students’ information behavior (search and evaluation of Web resources) to better understand a potential coupling between confirmatory search and polarization and, in further consequence, improve learning analytics and information services for individual and collective search in learning scenarios. In an empirical study, we had 91 high school students performing an information search in a social bookmarking environment. Gathered log data was used to compute indices of confirmatory search and polarisation as well as to analyze the impact of social stimulation. We find confirmatory search and polarization to correlate positively and social stimulation to mitigate, i.e., reduce the two variables’ relationship. From these findings, we derive practical implications for future work that aims to refine our formalism to compute confirmatory search and polarisation indices and to apply it for depolarizing information services
Kowald Dominik, Traub Matthias, Theiler Dieter, Gursch Heimo, Lacic Emanuel, Lindstaedt Stefanie , Kern Roman, Lex Elisabeth
2019
Kowald Dominik, Lacic Emanuel, Theiler Dieter, Traub Matthias, Kuffer Lucky, Lindstaedt Stefanie , Lex Elisabeth
2019
Kowald Dominik, Lex Elisabeth, Schedl Markus
2019
Lex Elisabeth, Kowald Dominik
2019
Fessl Angela, Kowald Dominik, Susana López Sola, Ana Moreno, Ricardo Alonso, Maturana, Thalmann_TU Stefan
2018
Learning analytics deals with tools and methods for analyzing anddetecting patterns in order to support learners while learning in formal as wellas informal learning settings. In this work, we present the results of two focusgroups in which the effects of a learning resource recommender system and adashboard based on analytics for everyday learning were discussed from twoperspectives: (1) knowledge workers as self-regulated everyday learners (i.e.,informal learning) and (2) teachers who serve as instructors for learners (i.e.,formal learning). Our findings show that the advantages of analytics for everydaylearning are three-fold: (1) it can enhance the motivation to learn, (2) it canmake learning easier and broadens the scope of learning, and (3) it helps to organizeand to systematize everyday learning.
Lex Elisabeth, Wagner Mario, Kowald Dominik
2018
In this work, we propose a content-based recommendation approach to increase exposure to opposing beliefs and opinions. Our aim is to help provide users with more diverse viewpoints on issues, which are discussed in partisan groups from different perspectives. Since due to the backfire effect, people's original beliefs tend to strengthen when challenged with counter evidence, we need to expose them to opposing viewpoints at the right time. The preliminary work presented here describes our first step into this direction. As illustrative showcase, we take the political debate on Twitter around the presidency of Donald Trump.
Kowald Dominik, Lex Elisabeth
2018
The micro-blogging platform Twitter allows its nearly 320 million monthly active users to build a network of follower connections to other Twitter users (i.e., followees) in order to subscribe to content posted by these users. With this feature, Twitter has become one of the most popular social networks on the Web and was also the first platform that offered the concept of hashtags. Hashtags are freely-chosen keywords, which start with the hash character, to annotate, categorize and contextualize Twitter posts (i.e., tweets).Although hashtags are widely accepted and used by the Twitter community, the heavy reuse of hashtags that are popular in the personal Twitter networks (i.e., own hashtags and hashtags used by followees) can lead to filter bubble effects and thus, to situations, in which only content associated with these hashtags are presented to the user. These filter bubble effects are also highly associated with the concept of confirmation bias, which is the tendency to favor and reuse information that confirms personal preferences. One example would be a Twitter user who is interested in political tweets of US president Donald Trump. Depending on the hashtags used, the user could either be stuck in a pro-Trump (e.g., #MAGA) or contra-Trump (e.g., #fakepresident) filter bubble. Therefore, the goal of this paper is to study confirmation bias and filter bubble effects in hashtag usage on Twitter by treating the reuse of hashtags as a phenomenon that fosters confirmation bias.
Lacic Emanuel, Kowald Dominik, Lex Elisabeth
2018
In this paper, we present work-in-progress on applying user pre-filtering to speed up and enhance recommendations based on Collab-orative Filtering. We propose to pre-filter users in order to extracta smaller set of candidate neighbors, who exhibit a high numberof overlapping entities and to compute the final user similaritiesbased on this set. To realize this, we exploit features of the high-performance search engine Apache Solr and integrate them into ascalable recommender system. We have evaluated our approachon a dataset gathered from Foursquare and our evaluation resultssuggest that our proposed user pre-filtering step can help to achieveboth a better runtime performance as well as an increase in overallrecommendation accuracy
Kowald Dominik, Lacic Emanuel, Theiler Dieter, Lex Elisabeth
2018
In this paper, we present preliminary results of AFEL-REC, a rec-ommender system for social learning environments. AFEL-RECis build upon a scalable so‰ware architecture to provide recom-mendations of learning resources in near real-time. Furthermore,AFEL-REC can cope with any kind of data that is present in sociallearning environments such as resource metadata, user interactionsor social tags. We provide a preliminary evaluation of three rec-ommendation use cases implemented in AFEL-REC and we €ndthat utilizing social data in form of tags is helpful for not only im-proving recommendation accuracy but also coverage. ‘is papershould be valuable for both researchers and practitioners inter-ested in providing resource recommendations in social learningenvironments
Dennerlein Sebastian, Kowald Dominik, Lex Elisabeth, Ley Tobias, Pammer-Schindler Viktoria
2018
Co-Creation methods for interactive computer systems design are by now widely accepted as part of the methodological repertoire in any software development process. As the communityis becoming more and more aware of the factthat software is driven by complex, artificially intelligent algorithms, the question arises what “co-creation of algorithms” in the sense of users ex-plicitly shaping the parameters of algorithms during co-creation, could mean, and how it would work. They are not tangible like featuresin a tool and desired effects are harder to be explained or understood. Therefore, we propose an it-erative simulation-based Co-Design approach that allows to Co-Create Algo-rithms together with the domain professionals by making their assumptions and effects observable. The proposal is a methodological idea for discussion within the EC-TEL community, yet to be applied in a research practice
Duricic Tomislav, Lacic Emanuel, Kowald Dominik, Lex Elisabeth
2018
User-based Collaborative Filtering (CF) is one of the most popularapproaches to create recommender systems. Œis approach is basedon €nding the most relevant k users from whose rating history wecan extract items to recommend. CF, however, su‚ers from datasparsity and the cold-start problem since users o‰en rate only asmall fraction of available items. One solution is to incorporateadditional information into the recommendation process such asexplicit trust scores that are assigned by users to others or implicittrust relationships that result from social connections betweenusers. Such relationships typically form a very sparse trust network,which can be utilized to generate recommendations for users basedon people they trust. In our work, we explore the use of a measurefrom network science, i.e. regular equivalence, applied to a trustnetwork to generate a similarity matrix that is used to select thek-nearest neighbors for recommending items. We evaluate ourapproach on Epinions and we €nd that we can outperform relatedmethods for tackling cold-start users in terms of recommendationaccuracy
Ross-Hellauer Anthony, Kowald Dominik, Lex Elisabeth
2018
d'Aquin Mathieu , Kowald Dominik, Fessl Angela, Thalmann Stefan, Lex Elisabeth
2018
The goal of AFEL is to develop, pilot and evaluate methods and applications, which advance informal/collective learning as it surfaces implicitly in online social environments. The project is following a multi-disciplinary, industry-driven approach to the analysis and understanding of learner data in order to personalize, accelerate and improve informal learning processes. Learning Analytics and Educational Data Mining traditionally relate to the analysis and exploration of data coming from learning environments, especially to understand learners' behaviours. However, studies have for a long time demonstrated that learning activities happen outside of formal educational platforms, also. This includes informal and collective learning usually associated, as a side effect, with other (social) environments and activities. Relying on real data from a commercially available platform, the aim of AFEL is to provide and validate the technological grounding and tools for exploiting learning analytics on such learning activities. This will be achieved in relation to cognitive models of learning and collaboration, which are necessary to the understanding of loosely defined learning processes in online social environments. Applying the skills available in the consortium to a concrete set of live, industrial online social environments, AFEL will tackle the main challenges of informal learning analytics through 1) developing the tools and techniques necessary to capture information about learning activities from (not necessarily educational) online social environments; 2) creating methods for the analysis of such informal learning data, based on combining feature engineering and visual analytics with cognitive models of learning and collaboration; and 3) demonstrating the potential of the approach in improving the understanding of informal learning, and the way it is better supported; 4) evaluate all the former items in real world large scale applications and platforms.
Kowald Dominik, Seitlinger Paul , Ley Tobias , Lex Elisabeth
2018
In this paper, we present the results of an online study with the aim to shed light on the impact that semantic context cues have on the user acceptance of tag recommendations. Therefore, we conducted a work-integrated social bookmarking scenario with 17 university employees in order to compare the user acceptance of a context-aware tag recommendation algorithm called 3Layers with the user acceptance of a simple popularity-based baseline. In this scenario, we validated and verified the hypothesis that semantic context cues have a higher impact on the user acceptance of tag recommendations in a collaborative tagging setting than in an individual tagging setting. With this paper, we contribute to the sparse line of research presenting online recommendation studies.
Lacic Emanuel, Kowald Dominik, Reiter-Haas Markus, Slawicek Valentin, Lex Elisabeth
2018
In this work, we address the problem of recommending jobs touniversity students. For this, we explore the impact of using itemembeddings for a content-based job recommendation system. Fur-thermore, we utilize a model from human memory theory to integratethe factors of frequency and recency of job posting interactions forcombining item embeddings. We evaluate our job recommendationsystem on a dataset of the Austrian student job portal Studo usingprediction accuracy, diversity as well as adapted novelty, which isintroduced in this work. We find that utilizing frequency and recencyof interactions with job postings for combining item embeddingsresults in a robust model with respect to accuracy and diversity, butalso provides the best adapted novelty results
d'Aquin Mathieu , Adamou Alessandro , Dietze Stefan , Fetahu Besnik , Gadiraju Ujwal , Hasani-Mavriqi Ilire, Holz Peter, Kümmerle Joachim, Kowald Dominik, Lex Elisabeth, Lopez Sola Susana, Mataran Ricardo, Sabol Vedran, Troullinou Pinelopi, Veas Eduardo, Veas Eduardo Enrique
2017
More and more learning activities take place online in a self-directed manner. Therefore, just as the idea of self-tracking activities for fitness purposes has gained momentum in the past few years, tools and methods for awareness and self-reflection on one's own online learning behavior appear as an emerging need for both formal and informal learners. Addressing this need is one of the key objectives of the AFEL (Analytics for Everyday Learning) project. In this paper, we discuss the different aspects of what needs to be put in place in order to enable awareness and self-reflection in online learning. We start by describing a scenario that guides the work done. We then investigate the theoretical, technical and support aspects that are required to enable this scenario, as well as the current state of the research in each aspect within the AFEL project. We conclude with a discussion of the ongoing plans from the project to develop learner-facing tools that enable awareness and self-reflection for online, self-directed learners. We also elucidate the need to establish further research programs on facets of self-tracking for learning that are necessarily going to emerge in the near future, especially regarding privacy and ethics.
Kowald Dominik, Lex Elisabeth
2017
In this paper, we study the imbalance between current state-of-the-art tag recommendation algorithms and the folksonomy structures of real-world social tagging systems. While algorithms such as FolkRank are designed for dense folksonomy structures, most social tagging systems exhibit a sparse nature. To overcome this imbalance, we show that cognitive-inspired algorithms, which model the tag vocabulary of a user in a cognitive-plausible way, can be helpful. Our present approach does this via implementing the activation equation of the cognitive architecture ACT-R, which determines the usefulness of units in human memory (e.g., tags). In this sense, our long-term research goal is to design hybrid recommendation approaches, which combine the advantages of both worlds in order to adapt to the current setting (i.e., sparse vs. dense ones)
Lacic Emanuel, Kowald Dominik, Lex Elisabeth
2017
Recommender systems are acknowledged as an essential instrumentto support users in finding relevant information. However,the adaptation of recommender systems to multiple domain-specificrequirements and data models still remains an open challenge. Inthe present paper, we contribute to this sparse line of research withguidance on how to design a customizable recommender systemthat accounts for multiple domains with heterogeneous data. Usingconcrete showcase examples, we demonstrate how to setup amulti-domain system on the item and system level, and we reportevaluation results for the domains of (i) LastFM, (ii) FourSquare,and (iii) MovieLens. We believe that our findings and guidelinescan support developers and researchers of recommender systemsto easily adapt and deploy a recommender system in distributedenvironments, as well as to develop and evaluate algorithms suitedfor multi-domain settings
Kowald Dominik, Kopeinik Simone , Lex Elisabeth
2017
Recommender systems have become important tools to supportusers in identifying relevant content in an overloaded informationspace. To ease the development of recommender systems, a numberof recommender frameworks have been proposed that serve a widerange of application domains. Our TagRec framework is one of thefew examples of an open-source framework tailored towards developingand evaluating tag-based recommender systems. In this paper,we present the current, updated state of TagRec, and we summarizeand reƒect on four use cases that have been implemented withTagRec: (i) tag recommendations, (ii) resource recommendations,(iii) recommendation evaluation, and (iv) hashtag recommendations.To date, TagRec served the development and/or evaluation processof tag-based recommender systems in two large scale Europeanresearch projects, which have been described in 17 research papers.‘us, we believe that this work is of interest for both researchersand practitioners of tag-based recommender systems.
Kowald Dominik, Pujari Suhbash Chandra, Lex Elisabeth
2017
Hashtags have become a powerful tool in social platformssuch as Twitter to categorize and search for content, and tospread short messages across members of the social network.In this paper, we study temporal hashtag usage practices inTwitter with the aim of designing a cognitive-inspired hashtagrecommendation algorithm we call BLLI,S. Our mainidea is to incorporate the effect of time on (i) individualhashtag reuse (i.e., reusing own hashtags), and (ii) socialhashtag reuse (i.e., reusing hashtags, which has been previouslyused by a followee) into a predictive model. For this,we turn to the Base-Level Learning (BLL) equation from thecognitive architecture ACT-R, which accounts for the timedependentdecay of item exposure in human memory. Wevalidate BLLI,S using two crawled Twitter datasets in twoevaluation scenarios. Firstly, only temporal usage patternsof past hashtag assignments are utilized and secondly, thesepatterns are combined with a content-based analysis of thecurrent tweet. In both evaluation scenarios, we find not onlythat temporal effects play an important role for both individualand social hashtag reuse but also that our BLLI,S approachprovides significantly better prediction accuracy andranking results than current state-of-the-art hashtag recommendationmethods.
Kowald Dominik, Lex Elisabeth, Kopeinik Simone
2016
In recent years, a number of recommendation algorithmshave been proposed to help learners find suitable learning resources online.Next to user-centered evaluations, offline-datasets have been usedto investigate new recommendation algorithms or variations of collaborativefiltering approaches. However, a more extensive study comparinga variety of recommendation strategies on multiple TEL datasets ismissing. In this work, we contribute with a data-driven study of recommendationstrategies in TEL to shed light on their suitability forTEL datasets. To that end, we evaluate six state-of-the-art recommendationalgorithms for tag and resource recommendations on six empiricaldatasets: a dataset from European Schoolnets TravelWell, a dataset fromthe MACE portal, which features access to meta-data-enriched learningresources from the field of architecture, two datasets from the socialbookmarking systems BibSonomy and CiteULike, a MOOC dataset fromthe KDD challenge 2015, and Aposdle, a small-scale workplace learningdataset. We highlight strengths and shortcomings of the discussed recommendationalgorithms and their applicability to the TEL datasets.Our results demonstrate that the performance of the algorithms stronglydepends on the properties and characteristics of the particular dataset.However, we also find a strong correlation between the average numberof users per resource and the algorithm performance. A tag recommenderevaluation experiment reveals that a hybrid combination of a cognitiveinspiredand a popularity-based approach consistently performs best onall TEL datasets we utilized in our study.
Traub Matthias, Lacic Emanuel, Kowald Dominik, Kahr Martin, Lex Elisabeth
2016
In this paper, we present work-in-progress on a recommender system designed to help people in need find the best suited social care institution for their personal issues. A key requirement in such a domain is to assure and to guarantee the person's privacy and anonymity in order to reduce inhibitions and to establish trust. We present how we aim to tackle this barely studied domain using a hybrid content-based recommendation approach. Our approach leverages three data sources containing textual content, namely (i) metadata from social care institutions, (ii) institution specific FAQs, and (iii) questions that a specific institution has already resolved. Additionally, our approach considers the time context of user questions as well as negative user feedback to previously provided recommendations. Finally, we demonstrate an application scenario of our recommender system in the form of a real-world Web system deployed in Austria.
Lacic Emanuel, Kowald Dominik, Lex Elisabeth
2016
Air travel is one of the most frequently used means of transportation in our every-day life. Thus, it is not surprising that an increasing number of travelers share their experiences with airlines and airports in form of online reviews on the Web. In this work, we thrive to explain and uncover the features of airline reviews that contribute most to traveler satisfaction. To that end, we examine reviews crawled from the Skytrax air travel review portal. Skytrax provides four review categories to review airports, lounges, airlines and seats. Each review category consists of several five-star ratings as well as free-text review content. In this paper, we conducted a comprehensive feature study and we find that not only five-star rating information such as airport queuing time and lounge comfort highly correlate with traveler satisfaction but also textual features in the form of the inferred review text sentiment. Based on our findings, we created classifiers to predict traveler satisfaction using the best performing rating features. Our results reveal that given our methodology, traveler satisfaction can be predicted with high accuracy. Additionally, we find that training a model on the sentiment of the review text provides a competitive alternative when no five star rating information is available. We believe that our work is of interest for researchers in the area of modeling and predicting user satisfaction based on available review data on the Web.
Kowald Dominik, Lex Elisabeth
2016
In this paper, we study factors that in uence tag reuse behavior in social tagging systems. Our work is guided by the activation equation of the cognitive model ACT-R, which states that the usefulness of information in human memory depends on the three factors usage frequency, recency and semantic context. It is our aim to shed light on the in uence of these factors on tag reuse. In our experiments, we utilize six datasets from the social tagging systems Flickr, CiteULike, BibSonomy, Delicious, LastFM and MovieLens, covering a range of various tagging settings. Our results con rm that frequency, recency and semantic context positively in uence the reuse probability of tags. However, the extent to which each factor individually in uences tag reuse strongly depends on the type of folksonomy present in a social tagging system. Our work can serve as guideline for researchers and developers of tag-based recommender systems when designing algorithms for social tagging environments.
Traub Matthias, Kowald Dominik, Lacic Emanuel, Lex Elisabeth, Schoen Pepjin, Supp Gernot
2015
In this paper, we present a scalable hotel recommender system for TripRebel, a new online booking portal. On the basis of the open-source enterprise search platform Apache Solr, we developed a system architecture with Web-based services to interact with indexed data at large scale as well as to provide hotel recommendations using various state-of-the-art recommender algorithms. We demonstrate the efficiency of our system directly using the live TripRebel portal where, in its current state, hotel alternatives for a given hotel are calculated based on data gathered from the Expedia AffiliateNetwork (EAN).
Dennerlein Sebastian, Kowald Dominik, Lex Elisabeth, Lacic Emanuel, Theiler Dieter, Ley Tobias
2015
Informal learning at the workplace includes a multitude of processes. Respective activities can be categorized into multiple perspectives on informal learning, such as reflection, sensemaking, help seeking and maturing of collective knowledge. Each perspective raises requirements with respect to the technical support, this is why an integrated solution relying on social, adaptive and semantic technologies is needed. In this paper, we present the Social Semantic Server, an extensible, open-source application server that equips clientside tools with services to support and scale informal learning at the workplace. More specifically, the Social Semantic Server semantically enriches social data that is created at the workplace in the context of user-to-user or user-artifact interactions. This enriched data can then in turn be exploited in informal learning scenarios to, e.g., foster help seeking by recommending collaborators, resources, or experts. Following the design-based research paradigm, the Social Semantic Server has been implemented based on design principles, which were derived from theories such as Distributed Cognition and Meaning Making. We illustrate the applicability and efficacy of the Social Semantic Server in the light of three real-world applications that have been developed using its social semantic services. Furthermore, we report preliminary results of two user studies that have been carried out recently.
Lacic Emanuel, Traub Matthias, Kowald Dominik, Lex Elisabeth
2015
In this paper, we present our approach towards an effective scalable recommender framework termed ScaR. Our framework is based on the microservices architecture and exploits search technology to provide real-time recommendations. Since it is our aim to create a system that can be used in a broad range of scenarios, we designed it to be capable of handling various data streams and sources. We show its efficacy and scalability with an initial experiment on how the framework can be used in a large-scale setting.
Lacic Emanuel, Luzhnica Granit, Simon Jörg Peter, Traub Matthias, Lex Elisabeth, Kowald Dominik
2015
In this paper, we present work-in-progress on a recommender system based on Collaborative Filtering that exploits location information gathered by indoor positioning systems. This approach allows us to provide recommendations for "extreme" cold-start users with absolutely no item interaction data available, where methods based on Matrix Factorization would not work. We simulate and evaluate our proposed system using data from the location-based FourSquare system and show that we can provide substantially better recommender accuracy results than a simple MostPopular baseline that is typically used when no interaction data is available.
Kowald Dominik, Lex Elisabeth
2015
To date, the evaluation of tag recommender algorithms has mostly been conducted in limited ways, including p-core pruned datasets, a small set of compared algorithms and solely based on recommender accuracy. In this study, we use an open-source evaluation framework to compare a rich set of state-of-the-art algorithms in six unfiltered, open datasets via various metrics, measuring not only accuracy but also the diversity, novelty and computational costs of the approaches. We therefore provide a transparent and reproducible tag recommender evaluation in real-world folksonomies. Our results suggest that the efficacy of an algorithm highly depends on the given needs and thus, they should be of interest to both researchers and developers in the field of tag-based recommender systems.
Kowald Dominik
2015
With the emergence of Web 2.0, tag recommenders have becomeimportant tools, which aim to support users in ndingdescriptive tags for their bookmarked resources. Althoughcurrent algorithms provide good results in terms of tag predictionaccuracy, they are often designed in a data-drivenway and thus, lack a thorough understanding of the cognitiveprocesses that play a role when people assign tags toresources. This thesis aims at modeling these cognitive dynamicsin social tagging in order to improve tag recommendationsand to better understand the underlying processes.As a rst attempt in this direction, we have implementedan interplay between individual micro-level (e.g., categorizingresources or temporal dynamics) and collective macrolevel(e.g., imitating other users' tags) processes in the formof a novel tag recommender algorithm. The preliminaryresults for datasets gathered from BibSonomy, CiteULikeand Delicious show that our proposed approach can outperformcurrent state-of-the-art algorithms, such as CollaborativeFiltering, FolkRank or Pairwise Interaction TensorFactorization. We conclude that recommender systems canbe improved by incorporating related principles of humancognition.
Seitlinger Paul, Kowald Dominik, Kopeinik Simone, Hasani-Mavriqi Ilire, Ley Tobias, Lex Elisabeth
2015
Classic resource recommenders like Collaborative Filtering(CF) treat users as being just another entity, neglecting non-linear user-resource dynamics shaping attention and inter-pretation. In this paper, we propose a novel hybrid rec-ommendation strategy that re nes CF by capturing thesedynamics. The evaluation results reveal that our approachsubstantially improves CF and, depending on the dataset,successfully competes with a computationally much moreexpensive Matrix Factorization variant.