@article{KorhonenWernerKutnaretal., author = {Korhonen, Elina and Werner, Andrea and Kutnar, Andreja and Toppinen, Anne and L{\"a}htinen, Katja}, title = {Communicating forest sector sustainability: results from four European countries}, series = {Forest Products Journal}, journal = {Forest Products Journal}, abstract = {Communication is an important tool in maintaining legitimacy and acceptability of forest sector operations and activities, and expectations by the general public on the forest sector conduct in Europe are in general very high. Despite this, there is scarce research in cross-national contexts on how forest sector sustainability is communicated to the general public and what development areas can be identified in terms of communication content. This study applies a qualitative content analysis in four forestry-rich European countries (Austria, Finland, Germany, and Slovenia). The state of online communication of 61 companies and 19 industry associations was qualitatively analyzed in 2014 with a focus on eight core sustainability topics of interest that were identified via an international forest sector stakeholder feedback process. Our results show some great similarities, but also some interesting differences in terms of communication frequency and weight of hot topics across countries. The most frequently communicated area was economic contribution of forests (in Finland and Austria), followed by debate over forest conservation versus production (Germany) and the concept-added value of wood (in Slovenia). With the exception of Slovenia, the role of forests in combating global warming was emphasized more frequently within industry associations than among individual forest industry companies. Characteristically, current content of sustainability communicatio n focuses on supplying factual information. Thus, there is a need for developing more targeted and bidirectional forms of stakeholder communication in the future, emphasizing also more active use of social media channels and empowering organizations to promote interactive communication and collaborative learning.}, subject = {Forest Science}, language = {en} } @article{ListSchwarzbauerBraunetal., author = {List, Julia and Schwarzbauer, Peter and Braun, Martin and Werner, Andrea and Langthaler, Georg and Stern, Tobias}, title = {Naive wood-supply predictions: Comparing two case studies from Austria}, series = {Austrian Journal Of Forest Science}, volume = {2016}, journal = {Austrian Journal Of Forest Science}, number = {2}, pages = {87 -- 110}, abstract = {Forest owner associations act as middlemen in the cooperative marketing of timber: they are supplied with small and fluctuating quantities of timber and sell bundled amounts to industrial consumers. Knowledge of the future quantity of monthly dis-tributable timber is of particular importance for planning, but remains a subject of uncertainty. This work presents models to predict wood supply based on a simple database. Models were tested in two case-study regions, which substantially differ Seite 88 J. List, P. Schwarzbauer, M. Braun, A. Werner, G. Langthaler, T. Sternin framework conditions for timber marketing. In each of the regions in Styria and Burgenland, different model types and subtypes were superior. It was concluded that models which determine timber supply in one forest association, are only restrictedly suitable to predict timber supply in another one.}, subject = {Forest Science}, language = {en} } @article{WoeberMehnenSykaceketal., author = {W{\"o}ber, Wilfried and Mehnen, Lars and Sykacek, Peter and Meimberg, Harald}, title = {Investigating Explanatory Factors of Machine Learning Models for Plant Classification}, series = {Plants}, volume = {2021}, journal = {Plants}, number = {10(12):2674}, pages = {20}, abstract = {Recent progress in machine learning and deep learning has enabled the implementation of plant and crop detection using systematic inspection of the leaf shapes and other morphological characters for identification systems for precision farming. However, the models used for this approach tend to become black-box models, in the sense that it is difficult to trace characters that are the base for the classification. The interpretability is therefore limited and the explanatory factors may not be based on reasonable visible characters. We investigate the explanatory factors of recent machine learning and deep learning models for plant classification tasks. Based on a Daucus carota and a Beta vulgaris image data set, we implement plant classification models and compare those models by their predictive performance as well as explainability. For comparison we implemented a feed forward convolutional neuronal network as a default model. To evaluate the performance, we trained an unsupervised Bayesian Gaussian process latent variable model as well as a convolutional autoencoder for feature extraction and rely on a support vector machine for classification. The explanatory factors of all models were extracted and analyzed. The experiments show, that feed forward convolutional neuronal networks (98.24\% and 96.10\% mean accuracy) outperforms the Bayesian Gaussian process latent variable pipeline (92.08\% and 94.31\% mean accuracy) as well as the convolutional autoenceoder pipeline (92.38\% and 93.28\% mean accuracy) based approaches in terms of classification accuracy, even though not significant for Beta vulgaris images. Additionally, we found that the neuronal network used biological uninterpretable image regions for the plant classification task. In contrast to that, the unsupervised learning models rely on explainable visual characters. We conclude that supervised convolutional neuronal networks must be used carefully to ensure biological interpretability. We recommend unsupervised machine learning, careful feature investigation, and statistical feature analysis for biological applications. View Full-Text Keywords: deep learning; machine learning; plant leaf morphometrics; explainable AI}, subject = {deep learning}, language = {en} }