Eine Zeitlang kannte die Para­meterzahl großer Sprachmodel­le nur eine Richtung: steil nach oben. Mehr Parameter bedingen mehr und hochwertigere Fähig­keiten, so die Überzeugung. Doch 2023 schlug die Stunde der mittelgroßen Sprach­KIs:  Sie sind genügsam – und  erstaunlich konkurrenzfähig. In mancher Disziplin rücken sie erstaunlich nahe an GPT-4 mit seinen kolportierten 1,8 Billionen Parametern heran. Damit tut sich ein riesiges Potenzial auf – auch für kleinere und mittelgroße Unternehmen, die mit eigenen  Anwendungen  liebäugeln.  Wir erklären, was die schlanken Verwandten der Giganten können, was sie so effizient macht und wie die Zukunft der Sprachmodelllandschaft aussehen könnte.

We present ASR Bundestag, a dataset for automatic speech recognition in German, consisting of 610 hours of aligned audio-transcript pairs for supervised training as well as 1,038 hours of unlabeled audio snippets for self-
supervised learning, based on raw audio data and transcriptions from plenary sessions and committee meetings of the German parliament. In addition, we discuss utilized approaches for the automated creation of speech datasets and assess the quality of the resulting dataset based on evaluations and finetuning of a pre-trained state of the art model. We make the dataset publicly available, including all subsets.

Large language models (LLMs) have garnered significant attention, but the definition of "large" lacks clarity. This paper focuses on medium-sized lan-guage models (MLMs), defined as having at least six billion parameters but less than 100 billion. The study evaluates MLMs regarding zero-shot genera-tive question answering, which requires models to provide elaborate answers without external document retrieval. The paper introduces an own test da-taset and presents results from human evaluation. Results show that combin-ing the best answers from different MLMs yielded an overall correct answer rate of 82.7% which is better than the 60.9% of ChatGPT. The best MLM achieved 46.4% and has 7B parameters, which highlights the importance of using appropriate training data for fine-tuning rather than solely relying on the number of parameters. More fine-grained feedback should be used to further improve the quality of answers.

The usage of e-learning platforms, online lectures and online meetings for academic teaching increased during the Covid-19 pandemic. Lecturers created video lectures, screencasts, or audio podcasts for online learning. The Hochschul-Assistenz-System (HAnS) is a learning experience platform that uses machine learning (ML) methods to support students and lecturers in the online learning and teaching processes. HAnS is being developed in multiple iterations as an agile open-source collaborative project supported by multiple universities and partners. This paper presents the current state of the development of HAnS on German video lectures.

The amount of freely available systems for automatic speech recognition (ASR) based on neural networks is growing steadily, with equally increasingly reliable predictions. However, the evaluation of trained models is typically exclusively based on statistical metrics such as WER or CER, which do not provide any insight into the nature or impact of the errors produced when predicting transcripts from speech input. This work presents a selection of ASR model architectures that are pretrained on the German language and evaluates them on a benchmark of diverse test datasets. It identifies cross-architectural prediction errors, classifies those into categories and traces the sources of errors per category back into training data as well as other sources. Finally, it discusses solutions in order to create qualitatively better training datasets and more robust ASR systems.

Autoregressive TTS models are still widely used. Due to their stochastic nature, the output may vary from very good to completely unusable from one inference to another. In this publication, we propose to use the percentage of completely correct transcribed sentences (PCTS) of an ASR system as a new objective quality measure for TTS inferences. PCTS is easy to measure and represents the intelligibility dimension of a typical subjective evaluation with mean opinion score (MOS). We show that PCTS leads to similar results as subjective MOS evaluation. A more detailed, semi-automatic error analysis of the differences between ASR transcripts of TTS speech and the text used for generating the TTS speech can help identifying problems in the TTS training data, that are harder to find with other methods.

Virtual reality (VR) is on the edge of getting a mainstream platform for gaming, education and product design. The feeling of being present in the virtual world is influenced by many factors and even more intriguing a single negative influence can destroy the illusion that was created with a lot of effort by other measures. Therefore, it is crucial to have a balance between the influencing factors, know the importance of the factors and have a good estimation of how much effort it takes to bring each factor to a certain level of fidelity. This paper collects influencing factors discussed in literature, analyses the immersion of current off-the-shelf VR-solutions and presents results from an empirical study on efforts and benefits from certain aspects influencing presence in VR experiences. It turns out, that sometimes delivering high fidelity is easier to achieve than medium fidelity and for other aspects it is worthwhile investing more effort to achieve higher fidelity to improve presence a lot.

While many speech synthesis systems based on deep neural networks are thoroughly evaluated and released for free use in English, models for languages with far less active speakers like German are scarcely trained and most often not published for common use. This work covers specific challenges in training text to speech models for the German language, including dataset selection and data preprocessing, and presents the training process for multiple models of an end-to-end text to speech system based on a combination of Tacotron 2 and Multi- Band MelGAN. All model compositions were evaluated against the mean opinion score, which revealed comparable results to models in literature that are trained and evaluated on English datasets. In addition, empirical analyses identified distinct aspects influencing the quality of such systems, based on subjective user experience. All trained models are released for public use.

Software integration, especially in Open Source Software (OSS), suffers from technical debt as software evolves over time. Evolvement often ends in incompatibility issues like changed or removed APIs. The problems increase, as more components need to be integrated. This article analyzes common problems in OSS integration with respect to technical debt esp. externally induced technical debt and collects empirical evidence from the case study "AMiProSI" which combines different functions of various OSS systems to create a unified intranet solution. As part of the analysis, this paper discusses component-based development and its related dependencies and explains related technical debt (TD). By bringing down the individual problems in sub-categories of technical debt, we show what kind of TD has arisen in our case study AMiProSI and give advice on how this could be prevented in other projects.

The increasing availability of audio data on the internet lead to a multitude of datasets for development and training of text to speech applications, based on neural networks. Highly differing quality of voice, low sampling rates, lack of text normalization and disadvantageous alignment of audio samples to corresponding transcript sentences still limit the performance of deep neural networks trained on this task. Additionally, data resources in languages like German are still very limited. We introduce the "HUI-Audio-Corpus-German", a large, open-source dataset for TTS engines, created with a processing pipeline, which produces high quality audio to transcription alignments and decreases manual effort needed for creation.

Deep learning has revived the field of automatic speech recognition (ASR) in the last ten years and pushed recognition rates into regions on par with humans. Applications like Siri, Amazon Alexa and Google Assistant are very popular, but have inherent privacy problems. In this paper, we evaluate state of the art open source ASR models regarding their usability in a smart speaker without cloud, both in terms of accuracy and runtime performance on cost-effective low power edge devices. We found Kaldi to be the most accurate solution and also among the fastest ones. It runs more than fast enough on an Nvidia Jetson Nano. It is still not on par with commercial cloud services, but getting close to it.

Deep learning is moving more and more from the cloud towards the edge. Therefore, embedded devices are needed that are reasonably cheap, energy-efficient and fast enough. In this paper we evaluate the performance and energy consumption of popular, off-the-shelf commercial devices for deep learning inferencing. We compare the Intel Neural Compute Stick 2, the Google Coral Edge TPU and the Nvidia Jetson Nano with the Raspberry Pi 4 for their suitability as a central controller in an autonomous vehicle for the formula student driverless.

Sprachsteuerung wurde mit Siri populär und ist dank Alexa dabei unseren Alltag weiter zu vereinfachen. Sprachassistenten und Chatbots können aber auch in Unternehmen selbst, oder in der Kommunikation mit deren Kunden gewinnbringend eingesetzt werden. Dank künstlicher Intelligenz wird die Qualität solcher Lösungen immer besser. Man sollte jedoch nicht den Versuchungen einer schnellen Cloud-Lösung erliegen, bei der der Datenschutz auf der Strecke bleibt.

Manufacturing execution systems (MES) are the central integration point for the shop floor. They collect information about customer orders from ERP systems (enterprise resource planning), calculate the best plans for production orders and their assignment to machines and monitor the execution of these plans. It is therefore also the central data hub on the shop floor and communicates progress back to the ERP system. However, each company and production facility is a bit different and it is a hard decision to find a good compromise between standard products with a long customization period, industry-specific solutions with less customization need and company-specific solutions with long development times. This paper proposes the use of business process management (BPM) as a means for easy graphical customization of production processes that lead to immediately executable workflows (zero code development) or need very few code additions to get executable (low code development).

Blockchain ist derzeit ein Hype. Bekanntlich folgt nach dem Gipfel der überzogenen Erwartungen der Absturz in das Tal der Enttäuschungen. Das liegt unter anderem daran, dass bisher nur wenige der sinnvollen Anwendungsfälle auch tatsächlich praktisch umgesetzt werden konnten und viele existierende Szenarien auch mit herkömmlichen Technologien ohne Blockchain funktionieren würden. Und dennoch kann Blockchain sinnvoll eingesetzt werden, beispielsweise auf dem Katasteramt oder in einer elektronischen Patientenakte.

IT support for collaboration has gone through quite some change since the beginning of the century and is providing more and more support for knowledge workers. Although single systems are getting easier to use, they are often not replacing former systems but accompany them, which makes the overall system landscape harder to oversee for knowledge workers. Future information systems should therefore combine the existing building blocks under a consistent user interface and assist the user in storing information at the right place. Seamlessly switching between formats, so that the user doesn’t have to decide upfront whether a blog entry, a wiki page or a text processor document is better suited for the information. This section discusses the development of digital collaboration solutions and shows how social software has changed them to better support knowledge processes.

Im folgenden Kapitel geht es um das Privatleben in der digitalen Welt. Wie wird es sich verändern? Welche Produkte werden wir nutzen? Was können sie heute schon, welche Funktionen und Fähigkeiten werden sie in naher Zukunft haben? Wird unser Leben dadurch angenehmer oder sollten wir uns vor so einer digitalen Zukunft fürchten? Anhand vieler Anwendungsbeispiele werden Sie verstehen, was das aktuelle Schlagwort Digitalisierung für das Privatleben bedeutet, und anhand zweier ausgewählter Technologien wird Ihnen ein kleiner Einblick hinter die Kulissen der technischen Entwicklung eröffnet.

Purpose – Technologies such as smart advisors, smart factories and the “Internet of things” have again sparked large interest in IT-enabled knowledge management. Although there have been numerous case studies and empirical studies on how the management of knowledge in large organizations has benefitted from these technologies, we have only scarce information on how small and medium-sized enterprises (SMEs) can manage the corresponding digital transformation necessary. In line with the emerging paradigm of physical-cyber-social computing (Sheth, Anantharam & Henson, 2013), we intend to foster the digital transformation that might boost SMEs’ resilience concerning the increasing demands of hypercompetitive environments. The research question discussed in this paper is how enterprise knowledge infrastructure services that are available publicly can support SMEs in building knowledge capabilities that strengthen their organizational resilience. Design/methodology/approach – The methodology used for this discussion consists of (1) a review of related work on the knowledge-based view, resilience and enterprise knowledge infrastructures plus (2) case examples taken from exploratory interviews with representatives of six German SMEs. As we study the relationships between IT, activities and outcomes, we build our research on the input-process-output model (McGrath, 1964) or, more specifically, an integrative framework for studying knowledge management enablers, processes, and organizational performance (Lee & Choi, 2003). Originality/value –Our contributions are two-fold: (1) we propose to employ capabilities in order to detail the concept of organizational resilience and (2) we present a model to illustrate how services of an enterprise knowledge infrastructure can support organizational resilience with the help of knowledge activities. Practical implications – The application of our approach to decompose organizational resilience into capabilities and to weave enterprise knowledge infrastructure services together with cloud service models to support such capabilities provides the basis for SMEs to benefit from comprehensive cloud-based offerings to build organizational resilience. By using such cloud offerings, SMEs can avoid issues with the development of own services. Cloud providers can use our decomposition of organizational resilience to design cloud-based IT services in a way that these services help to build capabilities for such resilience.