The InterConnect project gathers 50 European entities to develop and demonstrate advanced solutions for connecting and converging digital homes and buildings with the electricity sector. Machine Learning (ML) algorithms play a significant role in the InterConnect project. Most prominent are the services that do some kind of forecasting like predicting energy consumption for (Smart) devices and households in general. The SAREF ontology allows us to standardize input formats for common ML approaches and that explainability can be increased by selecting algorithms that inherently have these features (e.g. Decision Trees) and by using interactive web environments like Jupyter Notebooks a convenient solution for users is created where step by step the algorithmic procedures can be followed and visualized and forms an implementation example for explainable AI.
[This blog post is based on the Artificial Intelligence MSc thesis project from Fay Beening, supervised by myself and Joost de Boo, more information can be found on Fay’s website]
Recently, generative art has been one of the fields where AI, especially deep learning has caught the public eye. Algorithms and online tools such as Dall-E are able to produce astounding results based on large artistic datasets. One class of algorithms that has been at the root of this success is the Generative Adversarial Network (GAN), frequently used in online art-generating tools because of their ability to produce realistic artefacts.
but, is this “””real””” art? is this “””real””” creativity?
To address this, Fay investigated current theories on art and art education and found that these imply that true human creativity can be split into three types: 1) combinational, 2) explorative and 3) transformative creativity but that it also requires real-world experiences and interactions with people and the environment. Therefore, Fay in her thesis proposes to combine the GAN with an Internet of Things (IoT) setup to make it behave more creative.
Arduin-based prototype (image from Fay’s thesis)
She then designed a system that extends the original GAN with an interactive IoT system (implemented in an Arduino-based prototype) to simulate a more creative process. The prototype of the design showed a successful implementation of creative behaviour that can react to the environment and gradually change the direction of the generated images.
Images shown to the participant during the level of creativity task. Images 2 and 6 are creative GAN generated images. Images 1 and 5 are human-made art. Images 3 and 4 are online GAN generated art.
The generated art was evaluated based on their creativity by doing task-based interviews with domain experts. The results show that the the level to which the generated images are considered to be creative depends heavily on the participant’s view of creativity.
[This blog post was written by Nizar Hirzalla and describes his VU Master AI project conducted at the Koninklijke Bibliotheek (KB), co-supervised by Sara Veldhoen]
Authorship attribution is the process of correctly attributing a publication to its corresponding author, which is often done manually in real-life settings. This task becomes inefficient when there are many options to choose from due to authors having the same name. Authors can be defined by characteristics found in their associated publications, which could mean that machine learning can potentially automate this process. However, authorship attribution tasks introduce a typical class imbalance problem due to a vast number of possible labels in a supervised machine learning setting. To complicate this issue even more, we also use problematic data as input data as this mimics the type of available data for many institutions; data that is heterogeneous and sparse of nature.
The thesis searches for answers regarding how to automate authorship attribution with its known problems and this type of input data, and whether automation is possible in the first place. The thesis considers children’s literature and publications that can have between 5 and 20 potential authors (due to having the same exact name). We implement different types of machine learning methodologies for this method. In addition, we consider all available types of data (as provided by the National Library of the Netherlands), as well as the integration of contextual information.
Furthermore, we consider different types of computational representations for textual input (such as the title of the publication), in order to find the most effective representation for sparse text that can function as input for a machine learning model. These different types of experiments are preceded by a pipeline that consists out of pre-processing data, feature engineering and selection, converting data to other vector space representations and integrating linked data. This pipeline shows to actively improve performance when used with the heterogeneous data inputs.
Implemented neural network architectures for TFIDF (left) and Word2Vec (right) based text classification
Ultimately the thesis shows that automation can be achieved in up to 90% of the cases, and in a general sense can significantly reduce costs and time consumption for authorship attribution in a real-world setting and thus facilitate more efficient work procedures. While doing so, the thesis also finds the following key notions:
Between comparison of machine learning methodologies, two methodologies are considered: author classification and similarity learning. Author classification grants the best raw performance (F1. 0.92), but similarity learning provides the most robust predictions and increased explainability (F1. 0.88). For a real life setting with end users the latter is recommended as it presents a more suitable option for integration of machine learning with cataloguers, with only a small hit to performance.
The addition of contextual information actively increases performance, but performance depends on the type of information inclusion. Publication metadata and biographical author information are considered for this purpose. Publication metadata shows to have the best performance (predominantly the publisher and year of publication), while biographical author information in contrast negatively affects performance.
We consider BERT, word embeddings (Word2Vec and fastText) and TFIDF for representations of textual input. BERT ultimately grants the best performance; up to 200% performance increase when compared to word embeddings. BERT is a sophisticated language model with an applied transformer, which leads to more intricate semantic meaning representation of text that can be used to identify associated authors.
Based on surveys and interviews, we also find that end users mostly attribute importance to author related information when engaging in manual authorship attribution. Looking more in depth into the machine learning models, we can see that these primarily use publication metadata features to base predictions upon. We find that such differences in perception of information should ultimately not lead to negative experiences, as multiple options exist for harmonizing both parties’ usage of information.
Summary of the final performances of the best performing models from the differing implemented methodologies
