Relational and Semantic Data Mining

Lavrač, Nada; Vavpetič, Anže

doi:10.1007/978-3-319-23264-5_3

Cited by 11 publications

(6 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Finally, the authors introduced the Hedwig system [182], which overcomes some of the limitations of the previous systems. The findings of this series of work have been concluded in [183,184].…”

Section: Discussionmentioning

confidence: 81%

Semantic Web in Data Mining and Knowledge Discovery: A Comprehensive Survey

Ristoski

Paulheim

2016

SSRN Journal

View full text Add to dashboard Cite

a b s t r a c tData Mining and Knowledge Discovery in Databases (KDD) is a research field concerned with deriving higher-level insights from data. The tasks performed in that field are knowledge intensive and can often benefit from using additional knowledge from various sources. Therefore, many approaches have been proposed in this area that combine Semantic Web data with the data mining and knowledge discovery process. This survey article gives a comprehensive overview of those approaches in different stages of the knowledge discovery process. As an example, we show how Linked Open Data can be used at various stages for building content-based recommender systems. The survey shows that, while there are numerous interesting research works performed, the full potential of the Semantic Web and Linked Open Data for data mining and KDD is still to be unlocked.

show abstract

Section: Discussionmentioning

confidence: 81%

Semantic Web in Data Mining and Knowledge Discovery: A Comprehensive Survey

Ristoski

Paulheim

2016

SSRN Journal

View full text Add to dashboard Cite

show abstract

“…MRAM is the most recent approach which aims to overcome the difficulties in multi-relational data integration. It enables direct pattern extraction from multiple relations, without the necessity of transferring data to a single relation 25 28 , thus avoiding computationally expensive joining operations and semantic losses caused by the representation limit of a single table with repetitions of many attributes and data. Because this merged table is large and sparse, the mining process becomes more expensive and time-consuming 41 .…”

Section: Discussionmentioning

confidence: 99%

“…Three popular MRDM techniques are classification, clustering, and association. Association techniques (called multi-relational association mining - MRAM) have been successfully applied in bioinformatics, for example the analysis of gene set enrichment 28 , the prediction of hepatitis patients 29 , the analysis of different types of cancers based on microarray data 30 , the detection of potential adverse drug reactions 31 , and the prediction of protein interactions 32 . MRAM mines the data directly in their original structure of multiple relational tables, not requiring any pre-processing stage to generate a single table as in classical association mining (AM) algorithms like Apriori 33 and FP-growth 34 .…”

Section: Methodsmentioning

confidence: 99%

Novel drug target identification for the treatment of dementia using multi-relational association mining

Nguyễn

Priami

Caberlotto

2015

Sci Rep

View full text Add to dashboard Cite

Dementia is a neurodegenerative condition of the brain in which there is a progressive and permanent loss of cognitive and mental performance. Despite the fact that the number of people with dementia worldwide is steadily increasing and regardless of the advances in the molecular characterization of the disease, current medical treatments for dementia are purely symptomatic and hardly effective. We present a novel multi-relational association mining method that integrates the huge amount of scientific data accumulated in recent years to predict potential novel targets for innovative therapeutic treatment of dementia. Owing to the ability of processing large volumes of heterogeneous data, our method achieves a high performance and predicts numerous drug targets including several serine threonine kinase and a G-protein coupled receptor. The predicted drug targets are mainly functionally related to metabolism, cell surface receptor signaling pathways, immune response, apoptosis, and long-term memory. Among the highly represented kinase family and among the G-protein coupled receptors, DLG4 (PSD-95), and the bradikynin receptor 2 are highlighted also for their proposed role in memory and cognition, as described in previous studies. These novel putative targets hold promises for the development of novel therapeutic approaches for the treatment of dementia.

show abstract

“…The PILP setting was first introduced in [20], where three distinct settingsextended from traditional ILP [11] -are put forward: probabilistic entailment, probabilistic interpretations, and probabilistic proofs. Later, Raedt and Thon presented the system ProbFOIL [21], which is not only capable of performing induction over probabilistic examples, but also on background knowledge encoded as ProbLog probabilistic facts.…”

Section: Related Workmentioning

confidence: 99%

“…While researchers have spent their efforts on creating logic languages to represent probabilities and runtime environments that can deal with them [23,22,2,6,4,18], few works have been dedicated to learn rules from probabilistic knowledge. In this work, we introduce SkILL -a Stochastic Inductive Logic Learner -which can combine the rule learning capability of classic Inductive Logic Programming (ILP) [11,16] with uncertain knowledge as probabilistic annotated data to produce First Order Logic (FOL) theories.…”

Section: Introductionmentioning

confidence: 99%

SkILL - A Stochastic Inductive Logic Learner

Côrte-Real

Mantadelis

Dutra

et al. 2015

2015 IEEE 14th International Conference on Machine Learning and Applications (ICMLA)

View full text Add to dashboard Cite

Abstract. Probabilistic Inductive Logic Programming (PILP) is a relatively unexplored area of Statistical Relational Learning which extends classic Inductive Logic Programming (ILP). This work introduces SkILL, a Stochastic Inductive Logic Learner, which takes probabilistic annotated data and produces First Order Logic theories. Data in several domains such as medicine and bioinformatics have an inherent degree of uncertainty, that can be used to produce models closer to reality. SkILL can not only use this type of probabilistic data to extract non-trivial knowledge from databases, but it also addresses efficiency issues by introducing a novel, efficient and effective search strategy to guide the search in PILP environments. The capabilities of SkILL are demonstrated in three different datasets: (i) a synthetic toy example used to validate the system, (ii) a probabilistic adaptation of a well-known biological metabolism application, and (iii) a real world medical dataset in the breast cancer domain. Results show that SkILL can perform as well as a deterministic ILP learner, while also being able to incorporate probabilistic knowledge that would otherwise not be considered.

show abstract

Relational and Semantic Data Mining

Cited by 11 publications

References 22 publications

Semantic Web in Data Mining and Knowledge Discovery: A Comprehensive Survey

Semantic Web in Data Mining and Knowledge Discovery: A Comprehensive Survey

Novel drug target identification for the treatment of dementia using multi-relational association mining

SkILL - A Stochastic Inductive Logic Learner

Contact Info

Product

Resources

About