Review mining for music digital libraries

Current Music Information Retrieval (MIR) systems focus on melody based retrieval with some tolerance for user errors in the melody specification. The system described here presents a novel method for theme retrieval: A theme is described as a list of musical events, containing melody and harmony features, which must be presented in a given order and within a given time frame. The system retrieves musical phrases that fit the description. A system of this type could serve musicians and listeners who wish to discover thematically similar phrases in music digital libraries. The prototype and underlying model have been tested on midi sequences of music by W.A. Mozart and have shown good performance results.

Metadata research for music digital libraries has traditionally focused on genre. Despite its potential for improving the ability of users to better search and browse music collections, music subject metadata is an unexplored area. The objective of this study is to expand the scope of music metadata research, in particular, by exploring music subject classification based on user interpretations of music. Furthermore, we compare this previously unexplored form of user data to lyrics at subject prediction tasks. In our experiment, we use datasets consisting of 900 songs annotated with user interpretations. To determine the significance of performance differences between the two sources, we applied Friedman's ANOVA test on the classification accuracies. The results show that user-generated interpretations are significantly more useful than lyrics as classification features (p < 0.05). The findings support the possibility of exploiting various existing sources for subject metadata enrichment in music digital libraries.

The design and implementation of Harbin Institute of Technology--Digital Music Library (HIT-DML) is presented in this paper. Firstly, a novel framework, a music data model, and a query language are proposed as the theoretical foundation of the library. Secondly, music computing algorithms used in the library for feature extracting and matching are described. In addition, indices are introduced for both mining themes of music objects and accelerating content-based information retrieval. Finally, experimental results on the indices and the current development of the library are provided. HIT-DML is distinguished by the following points. First, it is inherently based on database systems, and combines database technologies with multimedia technologies seamlessly. Musical data are structurally stored. Second, it has a solid theoretical foundation, from framework and data model to query language. Last, it can retrieve musical information based on content against different kinds of musical instruments. The indices used, also power the library.

This paper describes work in progress on an ethnographic study of the searching/browsing techniques employed by people in the researcher’s local community. The insights provided by this type of study can inform the development of searching/browsing support for music digital libraries.

This poster describes a real-time audio-based automatic music genre classifier for use in organizing, browsing, and searching musical digital libraries. A decision tree classifier trained on a 40-dimension feature space is used to categorize music into one of 14 different genres with the results being displayed to a continuously updating user interface.

MiDiLiB is a six year research project on digital music libraries funded by the German Research Foundation (DFG) as a part of the Distributed Processing and Delivery of Digital Documents (V3D2) research initiative. MiDiLiB’s main focus is the development of content-based retrieval algorithms for both score- and waveform-based music. In this paper we give an overview of our research results, describe several prototypical systems for content-based music retrieval which have been developed during the project, and discuss applications of the presented techniques in the context of today’s and future digital music libraries.

This paper deals with the Music/Speech classification problem, starting from a set of features extracted directly from compressed audio data. The proposed classification system is able to label audio sequences stored as compressed MPEG layer III files. Decoding and analyzing in a unique stage is a fundamental tool for audio streaming applications, such as real time classification. Moreover, the techniques described herein provide useful tools in the management (data tagging, summarization, etc.) of a digital music library. The adopted set of short-time features are computed from the spectral information available in the decoding stage. In this paper, we show that for the classification problem at hand this set of features is redundant and can be dramatically pruned. To this aim we used an optimization strategy based on principal component analysis and genetic algorithms. The results show a very interesting classification accuracy using just one short-time feature.

Musicologists and musicians often would like to search by keys in a digital music library. In this paper, we introduce a new key-finding algorithm that can be applied to music in both symbolic and audio formats. The algorithm, which is based on a Hidden Markov Model (HMM), provides two stages of key-finding output; the first one referring to local keys and the second one to the global key. We describe the input, the two output stages, and the parameters of the model. In particular, we describe two configurable parameters, the transition probability distributions, which are based on a matrix of neighbouring keys, and the emission probability distributions, which make use of established key profiles. We discuss the local key-finding capabilities of the algorithm, presenting an example analysis of the Prelude Op. 28 No. 20 in C minor by Chopin, showing the local key regions obtained using different key profiles. We evaluate the global key-finding capabilities of the model, using an existing dataset and six well-known key profiles as different model parameters. Since different key profiles will tend to err or misclassify in different ways and across different pieces, we train an ensemble method with the predictions from all the key profiles (6) through our model. We show that the ensemble method achieves state-of-the-art performance for major and overall keys, however, it still underperforms the state-of-the-art for minor keys.

Content-Based Indexing of Spatial Objects in Digital Libraries

Digital libraries maintain information lifecycles for the digital objects contained therein. Typically these information lifecycles stipulate that a digital object is deleted at the end of its lifecycle. We propose a transient framework for digital libraries in which suitable objects are transformed through a state transition lifecycle thereby reducing the need to ever delete the digital object from the digital library, Transience works for certain media types and we describe and positively evaluate a prototype deployment in the domain of security video.

The FRBR-model provided by the IFLA Study group on the Functional Requirements for Bibliographic Records addresses the need for a more thorough model of bibliographic information. This paper describes a solution for applying the FRBR model to existing bibliographic catalogues. This is accomplished by augmenting the catalogue with an externally stored map that contains relationships and entities according to the model. The core of the system is the Digital Library LinkService developed at the Norwegian University of Science and Technology - a flexible and general purpose link service developed to support the structuring of information objects in digital libraries. A client application is developed to visualize the map and to enable users to interact with the map by navigating along the available paths.

Adaptability and interoperability in the field of highly interactive web-based courseware

This paper presents an approach of two-way data exchange between the citation content analysis, provided by the Cirtec project, and the big research digital library Socionet. Many papers in Socionet have citation relationships with other papers and also linkages with authors’ personal profiles and through them with other information objects. It allows making an enrichment of data for the citation content analysis by different additional information and, as well, linking results of such analysis with objects in a digital library, like papers, their authors, affiliation organizations, etc. We discuss what numeric and qualitative indicators can be built by citation content analysis based on the Cirtec open citation data. Since these indicators have IDs related with digital library objects, they can be integrated and visualized as computer-generated annotations to papers’ full texts in PDF.

While good practices are emerging with respect to publication of data alongside research outputs, we argue that computational descriptions (e.g. scripts, software and workflows) should also be included so that research can be interpreted, reconstructed and recomputed. A research article---or Research Object---should then describe all the components associated with a piece of digital research, including the descriptions of code and algorithms, effectively comprising an executable document. Furthermore we observe that such a re-executable object can be re-run automatically. The Music Information Retrieval research community has established community infrastructure and practices which are amenable to this approach, providing a glimpse of a future Music Digital Library. These ideas raise a number of issues for Digital Libraries more generally.

User reviews on gaming platforms such as Steam have become a crucial source of information for potential players before making purchasing decisions. Due to the varied nature of user opinions, sentiment analysis is essential for processing and interpreting these reviews. This study investigates the application of sentiment analysis to RPG game reviews on Steam, aiming to assist users by summarizing reviews through sentiment results and providing insights into the general perception of a game. To achieve this, the study applies sentiment analysis using Word2Vec and Support Vector Machine (SVM). It focuses on evaluating the impact of lemmatization during preprocessing and analyzing the performance of Word2Vec in sentiment classification. Word2Vec transforms review text into vector representations that capture semantic relationships, enhancing the model’s ability to understand context. Meanwhile, SVM is chosen as the classifier for its effectiveness in distinguishing between positive and negative reviews and handling high-dimensional data. The system developed uses Word2Vec with 300-dimensional vectors combined with an SVM Polynomial classifier, resulting in the best performance among the tested models. The final model achieves a macro-average F1-score of 88.6%, indicating a strong capability in accurately classifying sentiments in user reviews. These results highlight the potential of combining word embedding and machine learning techniques for analyzing sentiment in gaming platforms. Keywords: sentiment analysis, Word2Vec, SVM, Steam, RPG