Recommender systems are widely used in many areas, especially in e-commerce. Recently, they are a... more Recommender systems are widely used in many areas, especially in e-commerce. Recently, they are also applied in technology enhanced learning such as recommending resources (e.g. papers, books,...) to the learners (students). In this study, we propose using state-of-the-art recommender system techniques for predicting student performance. We introduce and formulate the problem of predicting student performance in the context of recommender systems. We present the matrix factorization methods, known as the most effective recommendation approaches, to implicitly take into account the latent factors, e.g. "slip" and "guess", in predicting student performance. Moreover, the knowledge of the learners has been improved over the time, thus, we propose tensor factorization methods to take the temporal effect into account. Experimental results show that the proposed approaches can improve the prediction results.
Predicting student performance (PSP) is the problem of predicting how well a student will perform... more Predicting student performance (PSP) is the problem of predicting how well a student will perform on a given task. It has gained more attention from the educational data mining community recently. Previous works show that good results can be achieved by casting the PSP to rating prediction problem in recommender systems, where students, tasks and performance scores are mapped to users, items and ratings respectively. One of the most prominent approaches for rating prediction which also performs well in PSP is matrix factorization (MF). However, the state-of-the-art MF approaches for PSP only make use of one relationship, that is, between students and tasks or students and skills needed to solve the tasks. In fact each student performs several tasks, and the tasks relate to the skill(s) needed to solve them, while students are also required mastering on the skills that they have learned. In this paper we propose to exploit such multiple relationships by using multi-relational MF methods. Experiments on three large datasets show that the proposed approach can improve the prediction results.
We tackle the problem of grouping content available in social media applications such as Flickr, ... more We tackle the problem of grouping content available in social media applications such as Flickr, Youtube, Panoramino etc. into clusters of documents describing the same event. This task has been referred to as event identification before. We present a new formalization of the event identification task as a record linkage problem and show that this formulation leads to a principled and highly efficient solution to the problem. We present results on two datasets derived from Flickr-last.fm and upcoming-comparing the results in terms of Normalized Mutual Information and F-Measure with respect to several baselines, showing that a record linkage approach outperforms all baselines as well as a stateof-the-art system. We demonstrate that our approach can scale to large amounts of data, reducing the processing time considerably compared to a state-of-theart approach. The scalability is achieved by applying an appropriate blocking strategy and relying on a Single Linkage clustering algorithm which avoids the exhaustive computation of pairwise similarities.
2009 Sixth International Conference on Information Technology New Generations, 2009
Exploring the metadata associated with documents in the semantic Web is a way to increase the pre... more Exploring the metadata associated with documents in the semantic Web is a way to increase the precision of information retrieval systems. Systems have been established so far failed to overcome fully the limitations of search based on keywords. Such systems are built from variations of classic models that represent information by keywords and work upon statistical correlations. This work proposes
A enorme quantidade de dados disponível na Web, fenômeno conhecido como sobrecarga de informação,... more A enorme quantidade de dados disponível na Web, fenômeno conhecido como sobrecarga de informação, cria uma demanda por aplicações que possam selecionar, de maneira automatizada, itens relevantes para os usuários. Para responder a essa demanda, os sistemas de recomendação utilizam algoritmos de filtragem de informação para suprir as necessidades de informação a longo prazo dos usuários. Entretanto, a falta de estruturação semântica dos dados disponíveis na Web compromete seriamente a eficiência de tais algoritmos. A Web Semântica apresenta-se como uma solução viável para este problema, visando estruturar a informação na Web de modo que esta possa ser compreendida e processada pelas entidades de software. É neste contexto que o InfoNorma, um sistema multiagente de recomendações para a área jurídica que se utiliza de tecnologias da Web Semântica, foi desenvolvido. A utilização de uma fonte baseada na Web tradicional acarretou em muitas dificuldades relativas ao processamento da fonte de informação. Tais dificuldades foram superadas graças ao uso de tecnologias da Web Semântica. Este artigo apresenta o InfoNorma e como o uso das tecnologias da Web Semântica contribuiu para a melhora da efetividade do sistema no acesso à informação.
Since ontology development is currently an error prone, time consuming and expensive task, ontolo... more Since ontology development is currently an error prone, time consuming and expensive task, ontology based systems suffer from the so called knowledge acquisition bottleneck. One approach for this problem is to provide automatic or semi-automatic support for ontology construction. This field of research is known as ontology learning. Many techniques for learning ontologies from text have been proposed, most of them based on statistical learning and natural language processing methods. This work presents an approach for extracting ontology concepts from text that combines both ideas through statistical relational learning. A Markov Logic Network as been developed for this task and is described here.
Ontologies constitute an approach for knowledge represen- tation that can be shared establishing ... more Ontologies constitute an approach for knowledge represen- tation that can be shared establishing a shared vocabulary for dierent applications and are also the backbone of the Semantic Web. Thus a fast and ecient ontology development is a requirement for the suc- cess of many knowledge based systems and for the Semantic Web itself. However, ontology development is a dicult and time consuming task. Ontology learning is an approach for the problem of knowledge acquisi- tion bottleneck that aims at reducing the cost of ontology construction through the development of automatic methods for the extraction of knowledge about a specific domain and its representation in an ontol- ogy like structure. This paper provides a discussion on existing ontology learning techniques and the state of the art of the field.
Proceedings of the fifth ACM international conference on Web search and data mining - WSDM '12, 2012
ABSTRACT A key element of the social networks on the internet such as Facebook and Flickr is that... more ABSTRACT A key element of the social networks on the internet such as Facebook and Flickr is that they encourage users to create connections between themselves, other users and objects. One important task that has been approached in the literature that deals with such data is to use social graphs to predict user behavior (e.g. joining a group of interest). More specifically, we study the cold-start problem, where users only participate in some relations, which we will call social relations, but not in the relation on which the predictions are made, which we will refer to as target relations. We propose a formalization of the problem and a principled approach to it based on multi-relational factorization techniques. Furthermore, we derive a principled feature extraction scheme from the social data to extract predictors for a classifier on the target relation. Experiments conducted on real world datasets show that our approach outperforms current methods.
Proceedings of the 27th Annual ACM Symposium on Applied Computing - SAC '12, 2012
ABSTRACT On RDF datasets, the truth values of triples are known when they are either explicitly s... more ABSTRACT On RDF datasets, the truth values of triples are known when they are either explicitly stated or can be inferred using logical entailment. Due to the open world semantics of RDF, nothing can be said about the truth values of triples that are neither in the dataset nor can be logically inferred. By estimating the truth values of such triples, one could discover new information from the database thus enabling to broaden the scope of queries to an RDF base that can be answered, support knowledge engineers in maintaining such knowledge bases or recommend users resources worth looking into for instance. In this paper, we present a new approach to predict the truth values of any RDF triple. Our approach uses a 3-dimensional tensor representation of the RDF knowledge base and applies tensor factorization techniques that take open world semantics into account to predict new true triples given already observed ones. We report results of experiments on real world datasets comparing different tensor factorization models. Our empirical results indicate that our approach is highly successful in estimating triple truth values on incomplete RDF datasets.
Recommender systems are widely used in many areas, especially in e-commerce. Recently, they are a... more Recommender systems are widely used in many areas, especially in e-commerce. Recently, they are also applied in technology enhanced learning such as recommending resources (e.g. papers, books,...) to the learners (students). In this study, we propose using state-of-the-art recommender system techniques for predicting student performance. We introduce and formulate the problem of predicting student performance in the context of recommender systems. We present the matrix factorization methods, known as the most effective recommendation approaches, to implicitly take into account the latent factors, e.g. "slip" and "guess", in predicting student performance. Moreover, the knowledge of the learners has been improved over the time, thus, we propose tensor factorization methods to take the temporal effect into account. Experimental results show that the proposed approaches can improve the prediction results.
Predicting student performance (PSP) is the problem of predicting how well a student will perform... more Predicting student performance (PSP) is the problem of predicting how well a student will perform on a given task. It has gained more attention from the educational data mining community recently. Previous works show that good results can be achieved by casting the PSP to rating prediction problem in recommender systems, where students, tasks and performance scores are mapped to users, items and ratings respectively. One of the most prominent approaches for rating prediction which also performs well in PSP is matrix factorization (MF). However, the state-of-the-art MF approaches for PSP only make use of one relationship, that is, between students and tasks or students and skills needed to solve the tasks. In fact each student performs several tasks, and the tasks relate to the skill(s) needed to solve them, while students are also required mastering on the skills that they have learned. In this paper we propose to exploit such multiple relationships by using multi-relational MF methods. Experiments on three large datasets show that the proposed approach can improve the prediction results.
We tackle the problem of grouping content available in social media applications such as Flickr, ... more We tackle the problem of grouping content available in social media applications such as Flickr, Youtube, Panoramino etc. into clusters of documents describing the same event. This task has been referred to as event identification before. We present a new formalization of the event identification task as a record linkage problem and show that this formulation leads to a principled and highly efficient solution to the problem. We present results on two datasets derived from Flickr-last.fm and upcoming-comparing the results in terms of Normalized Mutual Information and F-Measure with respect to several baselines, showing that a record linkage approach outperforms all baselines as well as a stateof-the-art system. We demonstrate that our approach can scale to large amounts of data, reducing the processing time considerably compared to a state-of-theart approach. The scalability is achieved by applying an appropriate blocking strategy and relying on a Single Linkage clustering algorithm which avoids the exhaustive computation of pairwise similarities.
2009 Sixth International Conference on Information Technology New Generations, 2009
Exploring the metadata associated with documents in the semantic Web is a way to increase the pre... more Exploring the metadata associated with documents in the semantic Web is a way to increase the precision of information retrieval systems. Systems have been established so far failed to overcome fully the limitations of search based on keywords. Such systems are built from variations of classic models that represent information by keywords and work upon statistical correlations. This work proposes
A enorme quantidade de dados disponível na Web, fenômeno conhecido como sobrecarga de informação,... more A enorme quantidade de dados disponível na Web, fenômeno conhecido como sobrecarga de informação, cria uma demanda por aplicações que possam selecionar, de maneira automatizada, itens relevantes para os usuários. Para responder a essa demanda, os sistemas de recomendação utilizam algoritmos de filtragem de informação para suprir as necessidades de informação a longo prazo dos usuários. Entretanto, a falta de estruturação semântica dos dados disponíveis na Web compromete seriamente a eficiência de tais algoritmos. A Web Semântica apresenta-se como uma solução viável para este problema, visando estruturar a informação na Web de modo que esta possa ser compreendida e processada pelas entidades de software. É neste contexto que o InfoNorma, um sistema multiagente de recomendações para a área jurídica que se utiliza de tecnologias da Web Semântica, foi desenvolvido. A utilização de uma fonte baseada na Web tradicional acarretou em muitas dificuldades relativas ao processamento da fonte de informação. Tais dificuldades foram superadas graças ao uso de tecnologias da Web Semântica. Este artigo apresenta o InfoNorma e como o uso das tecnologias da Web Semântica contribuiu para a melhora da efetividade do sistema no acesso à informação.
Since ontology development is currently an error prone, time consuming and expensive task, ontolo... more Since ontology development is currently an error prone, time consuming and expensive task, ontology based systems suffer from the so called knowledge acquisition bottleneck. One approach for this problem is to provide automatic or semi-automatic support for ontology construction. This field of research is known as ontology learning. Many techniques for learning ontologies from text have been proposed, most of them based on statistical learning and natural language processing methods. This work presents an approach for extracting ontology concepts from text that combines both ideas through statistical relational learning. A Markov Logic Network as been developed for this task and is described here.
Ontologies constitute an approach for knowledge represen- tation that can be shared establishing ... more Ontologies constitute an approach for knowledge represen- tation that can be shared establishing a shared vocabulary for dierent applications and are also the backbone of the Semantic Web. Thus a fast and ecient ontology development is a requirement for the suc- cess of many knowledge based systems and for the Semantic Web itself. However, ontology development is a dicult and time consuming task. Ontology learning is an approach for the problem of knowledge acquisi- tion bottleneck that aims at reducing the cost of ontology construction through the development of automatic methods for the extraction of knowledge about a specific domain and its representation in an ontol- ogy like structure. This paper provides a discussion on existing ontology learning techniques and the state of the art of the field.
Proceedings of the fifth ACM international conference on Web search and data mining - WSDM '12, 2012
ABSTRACT A key element of the social networks on the internet such as Facebook and Flickr is that... more ABSTRACT A key element of the social networks on the internet such as Facebook and Flickr is that they encourage users to create connections between themselves, other users and objects. One important task that has been approached in the literature that deals with such data is to use social graphs to predict user behavior (e.g. joining a group of interest). More specifically, we study the cold-start problem, where users only participate in some relations, which we will call social relations, but not in the relation on which the predictions are made, which we will refer to as target relations. We propose a formalization of the problem and a principled approach to it based on multi-relational factorization techniques. Furthermore, we derive a principled feature extraction scheme from the social data to extract predictors for a classifier on the target relation. Experiments conducted on real world datasets show that our approach outperforms current methods.
Proceedings of the 27th Annual ACM Symposium on Applied Computing - SAC '12, 2012
ABSTRACT On RDF datasets, the truth values of triples are known when they are either explicitly s... more ABSTRACT On RDF datasets, the truth values of triples are known when they are either explicitly stated or can be inferred using logical entailment. Due to the open world semantics of RDF, nothing can be said about the truth values of triples that are neither in the dataset nor can be logically inferred. By estimating the truth values of such triples, one could discover new information from the database thus enabling to broaden the scope of queries to an RDF base that can be answered, support knowledge engineers in maintaining such knowledge bases or recommend users resources worth looking into for instance. In this paper, we present a new approach to predict the truth values of any RDF triple. Our approach uses a 3-dimensional tensor representation of the RDF knowledge base and applies tensor factorization techniques that take open world semantics into account to predict new true triples given already observed ones. We report results of experiments on real world datasets comparing different tensor factorization models. Our empirical results indicate that our approach is highly successful in estimating triple truth values on incomplete RDF datasets.
Uploads
Papers by Lucas Drumond