Information extraction for knowledge base construction in the music domain

Oramas, Sergio; Espinosa-Anke, Luis; Sordo, Mohamed; Saggion, Horacio; Serra, Xavier

doi:10.1016/j.datak.2016.06.001

Cited by 32 publications

(27 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Recently, Oramas et al [34] proposed a new method to build a music knowledge base entirely from scratch and they extracted relationships between pairs of entities by combining a state-of-theart linking tool with a rule-based algorithm. The result of this process revealed a number of facts that were not known in common knowledge repositories (e.g., Wikipedia).…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

Generating and Understanding Personalized Explanations in Hybrid Recommender Systems

Kouki¹,

Schaffer

Pujara

et al. 2020

ACM Trans. Interact. Intell. Syst.

View full text Add to dashboard Cite

Recommender systems are ubiquitous and shape the way users access information and make decisions. As these systems become more complex, there is a growing need for transparency and interpretability. In this article, we study the problem of generating and visualizing personalized explanations for recommender systems that incorporate signals from many different data sources. We use a flexible, extendable probabilistic programming approach and show how we can generate real-time personalized recommendations. We then turn these personalized recommendations into explanations. We perform an extensive user study to evaluate the benefits of explanations for hybrid recommender systems. We conduct a crowd-sourced user study where our system generates personalized recommendations and explanations for real users of the last.fm music platform. First, we evaluate the performance of the recommendations in terms of perceived accuracy and novelty. Next, we experiment with (1) different explanation styles (e.g., user-based, item-based), (2) manipulating the number of explanation styles presented, and (3) manipulating the presentation format (e.g., textual vs. visual). We also apply a mixed-model statistical analysis to consider user personality traits as a control variable and demonstrate the usefulness of our approach in creating personalized hybrid explanations with different style, number, and format. Finally, we perform a post analysis that shows different preferences for explanation styles between experienced and novice last.fm users.

show abstract

Section: Related Workmentioning

confidence: 99%

“…To summarize, our work differs from prior work in the following ways. First, existing work on explanations either does not involve a recommendation algorithm [5,29] or uses a baseline recommender [7,16,34,34]. As a result, comparing the accuracy of the recommendations of these approaches with other algorithms is prohibitive.…”

Section: Related Workmentioning

confidence: 99%

Generating and Understanding Personalized Explanations in Hybrid Recommender Systems

Kouki¹,

Schaffer

Pujara

et al. 2020

ACM Trans. Interact. Intell. Syst.

View full text Add to dashboard Cite

show abstract

“…But these methods need annotated corpora to train the model to identify pair of entities. To overcome this limitation, unsupervised and semi-supervised methods were introduced that uses heuristic rules or different type of clustering algorithms to identify relations between entities from large unlabeled corpus [11], [12]. Several hybrid approaches have also been introduced to maximize the efficiency for different languages, but linguistic components and domain criteria affects the efficiency and accuracy of these approaches.…”

Section: A Information Extraction From Text Datamentioning

confidence: 99%

Information Extraction from Multifaceted Unstructured Big Data

2019

IJRTE

View full text Add to dashboard Cite

In the era of digital globalization, huge volume and variety of data are being produced at a very high rate. Every day, the world is producing around 2.5 quintillion bytes of data. According to IDC, by 2020, over 40 zettabytes of data will be generated and reproduced. Digital data have become a deluge, overwhelming in every field of information technology (IT), business, science and engineering. These fields are shifting to smart and advanced technologies such as smart manufacturing industries, data-aware medical sciences, and other smart applications. These applications are facilitating the industries in context of data-driven decision making, big data storage, and complex analysis of large data sets. Also, these applications are contributing to generate big data deluge where a variety of data necessitate the industries to use advanced IT approaches. 95% of the digital universe is unstructured data. It is rich data as it contains information that can play a vital role to improve big data analytics. The heterogeneity, complexity, lack of structured information, poor quality and scalability of unstructured data generates difficulties in adapting traditional information extraction techniques. Information extraction can play a vital role in transformation of unstructured data into useful information. A multistep pipeline with data preprocessing steps, extraction methods and representation are utmost requirement to improve the unstructured data analytics. In this regard, this paper presents a short review of information extraction process w.r.t. input data type, extraction methods with their corresponding techniques, and representation of extracted information. The issues with unstructured data and the challenges to information extraction from multifaceted unstructured big data as well as the future research directions have also been discussed

show abstract

“…EL systems are useful for music recommendation [28]. However, they are not optimized for the music domain, and are prone to errors [27]. e application of a ltering process over the set of identi ed entities based on their classi cation within the DBpedia Ontology, has demonstrated its utility to improve music retrieval tasks, such as artist similarity [29].…”

Section: Semantic Enrichmentmentioning

confidence: 99%

A Deep Multimodal Approach for Cold-start Music Recommendation

Oramas

Nieto

Sordo

et al. 2017

Proceedings of the 2nd Workshop on Deep Learning for Recommender Systems

Self Cite

View full text Add to dashboard Cite

An increasing amount of digital music is being published daily. Music streaming services o en ingest all available music, but this poses a challenge: how to recommend new artists for which prior knowledge is scarce? In this work we aim to address this so-called cold-start problem by combining text and audio information with user feedback data using deep network architectures. Our method is divided into three steps. First, artist embeddings are learned from biographies by combining semantics, text features, and aggregated usage data. Second, track embeddings are learned from the audio signal and available feedback data. Finally, artist and track embeddings are combined in a multimodal network. Results suggest that both spli ing the recommendation problem between feature levels (i.e., artist metadata and audio track), and merging feature embeddings in a multimodal approach improve the accuracy of the recommendations.

show abstract

Information extraction for knowledge base construction in the music domain

Cited by 32 publications

References 24 publications

Generating and Understanding Personalized Explanations in Hybrid Recommender Systems

Generating and Understanding Personalized Explanations in Hybrid Recommender Systems

Information Extraction from Multifaceted Unstructured Big Data

A Deep Multimodal Approach for Cold-start Music Recommendation

Contact Info

Product

Resources

About