Franziska Horn

Franziska Horn

With an exponentially growing number of scientific papers published each year, advanced tools for exploring and discovering publications of interest are becoming indispensable. To empower users beyond a simple keyword search provided e.g. by Google Scholar, we present the novel web application PubVis. Powered by a variety of machine learning techniques, it combines essential features to help researchers find the content most relevant to them. An interactive visualization of a large collection of scientific publications provides an overview of the field and encourages the user to explore articles beyond a narrow research focus. This is augmented by personalized content based article recommendations as well as an advanced full text search to discover relevant references. The open sourced implementation of the app can be easily set up and run locally on a desktop computer to provide access to content tailored to the specific needs of individual users. Additionally, a PubVis demo with access to a collection of 10,000 papers can be tested online.

Mihail Bogojeski, Simeon Sauer, Franziska Horn et al.

Accurately predicting industrial aging processes makes it possible to schedule maintenance events further in advance, ensuring a cost-efficient and reliable operation of the plant. So far, these degradation processes were usually described by mechanistic or simple empirical prediction models. In this paper, we evaluate a wider range of data-driven models, comparing some traditional stateless models (linear and kernel ridge regression, feed-forward neural networks) to more complex recurrent neural networks (echo state networks and LSTMs). We first examine how much historical data is needed to train each of the models on a synthetic dataset with known dynamics. Next, the models are tested on real-world data from a large scale chemical plant. Our results show that recurrent models produce near perfect predictions when trained on larger datasets, and maintain a good performance even when trained on smaller datasets with domain shifts, while the simpler models only performed comparably on the smaller datasets.

Franziska Horn, Robert Pack, Michael Rieger

This paper describes the autofeat Python library, which provides scikit-learn style linear regression and classification models with automated feature engineering and selection capabilities. Complex non-linear machine learning models, such as neural networks, are in practice often difficult to train and even harder to explain to non-statisticians, who require transparent analysis results as a basis for important business decisions. While linear models are efficient and intuitive, they generally provide lower prediction accuracies. Our library provides a multi-step feature engineering and selection process, where first a large pool of non-linear features is generated, from which then a small and robust set of meaningful features is selected, which improve the prediction accuracy of a linear model while retaining its interpretability.

Lea Helmers, Franziska Horn, Franziska Biegler et al.

More than ever, technical inventions are the symbol of our society's advance. Patents guarantee their creators protection against infringement. For an invention being patentable, its novelty and inventiveness have to be assessed. Therefore, a search for published work that describes similar inventions to a given patent application needs to be performed. Currently, this so-called search for prior art is executed with semi-automatically composed keyword queries, which is not only time consuming, but also prone to errors. In particular, errors may systematically arise by the fact that different keywords for the same technical concepts may exist across disciplines. In this paper, a novel approach is proposed, where the full text of a given patent application is compared to existing patents using machine learning and natural language processing techniques to automatically detect inventions that are similar to the one described in the submitted document. Various state-of-the-art approaches for feature extraction and document comparison are evaluated. In addition to that, the quality of the current search process is assessed based on ratings of a domain expert. The evaluation results show that our automated approach, besides accelerating the search process, also improves the search results for prior art with respect to their quality.

Robert Greinacher, Franziska Horn

Producing the required amounts of training data for machine learning and NLP tasks often involves human annotators doing very repetitive and monotonous work. In this paper, we present and evaluate our novel annotation framework DALPHI, which facilitates the annotation process by providing the annotator with suggestions generated by an automated, active-learning based assistance system. In a study with 66 participants, we demonstrate on the exemplary task of annotating named entities in text documents that with this assistance system the annotation processes can be improved with respect to the quality and quantity of produced annotations, even if the pre-annotations provided by the assistance system are at a recall level of only 50%.

Franziska Horn, Leila Arras, Grégoire Montavon et al.

When dealing with large collections of documents, it is imperative to quickly get an overview of the texts' contents. In this paper we show how this can be achieved by using a clustering algorithm to identify topics in the dataset and then selecting and visualizing relevant words, which distinguish a group of documents from the rest of the texts, to summarize the contents of the documents belonging to each topic. We demonstrate our approach by discovering trending topics in a collection of New York Times article snippets.

Franziska Horn, Leila Arras, Grégoire Montavon et al.

When working with a new dataset, it is important to first explore and familiarize oneself with it, before applying any advanced machine learning algorithms. However, to the best of our knowledge, no tools exist that quickly and reliably give insight into the contents of a selection of documents with respect to what distinguishes them from other documents belonging to different categories. In this paper we propose to extract `relevant words' from a collection of texts, which summarize the contents of documents belonging to a certain class (or discovered cluster in the case of unlabeled datasets), and visualize them in word clouds to allow for a survey of salient features at a glance. We compare three methods for extracting relevant words and demonstrate the usefulness of the resulting word clouds by providing an overview of the classes contained in a dataset of scientific publications as well as by discovering trending topics from recent New York Times article snippets.

Franziska Horn

With a simple architecture and the ability to learn meaningful word embeddings efficiently from texts containing billions of words, word2vec remains one of the most popular neural language models used today. However, as only a single embedding is learned for every word in the vocabulary, the model fails to optimally represent words with multiple meanings. Additionally, it is not possible to create embeddings for new (out-of-vocabulary) words on the spot. Based on an intuitive interpretation of the continuous bag-of-words (CBOW) word2vec model's negative sampling training objective in terms of predicting context based similarities, we motivate an extension of the model we call context encoders (ConEc). By multiplying the matrix of trained word2vec embeddings with a word's average context vector, out-of-vocabulary (OOV) embeddings and representations for a word with multiple meanings can be created based on the word's local contexts. The benefits of this approach are illustrated by using these word embeddings as features in the CoNLL 2003 named entity recognition (NER) task.

Franziska Horn, Klaus-Robert Müller

Matrix factorization is at the heart of many machine learning algorithms, for example, dimensionality reduction (e.g. kernel PCA) or recommender systems relying on collaborative filtering. Understanding a singular value decomposition (SVD) of a matrix as a neural network optimization problem enables us to decompose large matrices efficiently while dealing naturally with missing values in the given matrix. But most importantly, it allows us to learn the connection between data points' feature vectors and the matrix containing information about their pairwise relations. In this paper we introduce a novel neural network architecture termed Similarity Encoder (SimEc), which is designed to simultaneously factorize a given target matrix while also learning the mapping to project the data points' feature vectors into a similarity preserving embedding space. This makes it possible to, for example, easily compute out-of-sample solutions for new data points. Additionally, we demonstrate that SimEc can preserve non-metric similarities and even predict multiple pairwise relations between data points at once.

Leila Arras, Franziska Horn, Grégoire Montavon et al.

Text documents can be described by a number of abstract concepts such as semantic category, writing style, or sentiment. Machine learning (ML) models have been trained to automatically map documents to these abstract concepts, allowing to annotate very large text collections, more than could be processed by a human in a lifetime. Besides predicting the text's category very accurately, it is also highly desirable to understand how and why the categorization process takes place. In this paper, we demonstrate that such understanding can be achieved by tracing the classification decision back to individual words using layer-wise relevance propagation (LRP), a recently developed technique for explaining predictions of complex non-linear classifiers. We train two word-based ML models, a convolutional neural network (CNN) and a bag-of-words SVM classifier, on a topic categorization task and adapt the LRP method to decompose the predictions of these models onto words. Resulting scores indicate how much individual words contribute to the overall classification decision. This enables one to distill relevant information from text documents without an explicit semantic information extraction step. We further use the word-wise relevance scores for generating novel vector-based document representations which capture semantic information. Based on these document vectors, we introduce a measure of model explanatory power and show that, although the SVM and CNN models perform similarly in terms of classification accuracy, the latter exhibits a higher level of explainability which makes it more comprehensible for humans and potentially more useful for other applications.

Leila Arras, Franziska Horn, Grégoire Montavon et al.

Layer-wise relevance propagation (LRP) is a recently proposed technique for explaining predictions of complex non-linear classifiers in terms of input variables. In this paper, we apply LRP for the first time to natural language processing (NLP). More precisely, we use it to explain the predictions of a convolutional neural network (CNN) trained on a topic categorization task. Our analysis highlights which words are relevant for a specific prediction of the CNN. We compare our technique to standard sensitivity analysis, both qualitatively and quantitatively, using a "word deleting" perturbation experiment, a PCA analysis, and various visualizations. All experiments validate the suitability of LRP for explaining the CNN predictions, which is also in line with results reported in recent image classification studies.