Analytical Reasoning by applying machine learning approaches, artificial intelligence, NLP and visualizations allow to get deep insights into the different domains of various stakeholders and enable to solve complex tasks. Thereby the tasks are very heterogenous and subject of investigation in the different areas of application. These tasks or challenges should be defined by the stakeholders themselves and lead through a deep investigation to advanced analytical approaches. We therefore set up a strategic alliance of research, enterprises and societal organization with the goal of a strong collaboration to identify in a first step these challenges and workout technological solutions for each application scenario. We give in this paper a first draft of current challenges and technological advancements. The main contribution of this paper is next to an accurate description of the current challenges in the analytics domain, also the description of an agenda how these challenges can be solved. Furthermore, a process is explained, how the strategic alliance should act and organize their work to realize beneficial and useful analytical solutions.
Analytical Reasoning by applying machine learning approaches, artificial intelligence, NLP and visualizations allow to get deep insights into the different domains of various stakeholders and enable to solve complex tasks. An example for such a complex task could be the identification and prediction of new technologies, upcoming technological and methodical trends and innovations in the different domains of application. This is essential for making strategic decisions for economical and societal challenges. The analytical approaches can be supported through trend mining, machine learning, artificial intelligence, visual analytics and simulation.
To face the named aspects, a strategic alliance is necessary that seriously considers a variety of data and focus in particular on huge amount of data, streaming data and unexplored free data from different resources. But even more, the alliance should not
Business and Society 2 have only the data as basement in mind, it is much more important to follow the entire data processing pipeline. So that also data storing, data processing and data visualization needs to adequately handled for an effective and efficient analysis.
The main goal is to optimize and use technological and methodological innovations to identify required analytical approaches for politics, business, education and research and to develop adequate solutions to the identified challenges. Therefore, an interdisciplinary European network should be initiated that relies on the expertise of the involved partners, who are internationally renowned in their respective fields. From a technological point of view, methods of digitization, artificial intelligence, natural language processing, visual analytics, data analytics and simulation will be optimized and used for different application scenarios. Thereby, the actors in the respective areas define the challenges themselves, whereby data-driven trends will also be identified. The validation of the results is also carried out by the actors of the respective areas. Therefore, a network of interdisciplinary partners has been established who are able to qualitatively define and assess those challenges and solutions. In this paper we describe the major challenges, particularly in Europe, how the network aims to face these challenges and finally how the network actions will be organized. 2
The establishment of analytics and analytical solution is a challenge in most countries. This challenge is often aligned to the digitalization challenge, since wide ranges of current business are still analog. But to face nowadays developments, it is necessary to consider enhancements of products and services to have the address the changed needs of customers.
However, the challenges in using and establishing analytics in daily business and processes requires a more precise understanding. In this section, the intention is to outline a variety of challenges that not only cover technological aspects, but also societal and economical challenges. 2.1
Recent advances in technologies and especially artificial intelligence (AI) thrive whole
new business concepts in digitizing current labor or is the basis of completely new
industries. In the Global Risk Report 2019, the WEF identified the “adverse consequences
of technological advances” as one of the major risks [
Due to the rise of a variety of machine learning, Artificial Intelligence, data mining
and visual analytics methods, it is necessary and important to evaluate the potentials of
all kinds of technologies [
The WEF identified “unemployment or underemployment” as one of the many major
global risks in their Global Risk Report 2019 [
The Global Risk Report additionally identified the “spread of infectious diseases” as
one of the major risks that we are currently facing [
In an ever more complex world, which develops at an ever-faster pace, governments
are struggling to keep up and to address societal problems in a sound and timely
manner. This led to a decline in acceptance of government decisions throughout the
democratic world, leading to a rise of citizens and associations who are critically questioning
democratic governments to the extent that some groups of people are willing to
terminate the social contract on their side. To countermeasure this trend societies place their
hopes on e-governance and new forms of policy modelling giving stakeholders, from
scientists to business associations over civil society groups down to the single
individual possibilities to access information and represent their insights and interests in the
decision-making processes. However, current policy making is still a highly analog
process and technologies are still barely used or not implemented consequently [
More and more data are getting digitized and allows new insights into market,
technologies, competitors and more [
Analytics is a way to simplify the assessment of complex data and has the potential to
enable, both, the private sector and society to assess, understand and process
unstructured data to develop new strategies, in areas, such as corporate foresight and the future
of employment with regard to their challenges [
The analysis process requires a number of different steps to allow the application of
analytical technologies, e.g. artificial intelligence, machine learning and simulation. It
is necessary to investigate the entire digital data transformation, e.g. integrate data,
extract information from the integrated data and provide effective and interactive analysis
tools. Scientific open access data is maybe one resource that the European Commission
will probably address in the new Framework as the current suggestions imply [
Given data in the web is often incomplete or not well defined. In consequence it is still
difficult to identify a given entity in a full-automated and sufficient way. Also, the
categorization of data is often not given. To bridge this knowledge gap, data mining
techniques are suitable approaches to mine such information with regard to accessible data
[
Besides the pre-processing of data, data mining, simulation, foresight and prediction
approaches will enable to evaluate the underlying methods and enable considering
different approaches of machine learning and artificial intelligence to identify the
emerging technologies and trends. It is necessary to evaluate the approaches from the
business-perspective [
For trend identification as an example for analytics, an encompassing view on the data
is essential to “see” upcoming or manifesting trends [
Visual analytics helps to view complex information in a simplified manner. Depicting the smart manufacturing example, a lot of technologies are awaiting at the edge, that will disrupt the training, the daily work and the processes within the domain. Trainees have to understand more and more complex cyber-physical systems that are also interconnected and fully-automated. This machinery produces a lot of complex information that is difficult to assess and therewith processes and procedures are difficult to follow and learn from. Besides the training process, the daily work as surveillance and maintenance is also directly affected by this complexity, followed by the internal factory processes. Visual Analytics can help process, annotate, identify errors, visualize dependencies and connections to accelerate training, daily work or the internal factory processes by mitigating complexity and lower the entropy to make systems and processes more accessible. Therefore, Visual Analytics helps trainees and professionals to stick with time restraints without letting the learning curve become too steep with all the new technologies. Consequently, leaving nobody behind by technological advances.
The strategic alliance owns a variety of technologies and expertise in the named fields, however, many of these technologies focus on only a few aspects, yet. Within the action these technologies and expertise should be merged to advance the power and effectiveness and enable a broader usage. The technological outcome will finally lead to new scientific solutions that are fruitful for a wide number of societal and economical use cases, e.g. “visual analytics for policy modelling", "visual analytics for medical purposes", "visual analytics for smart businesses" or - to frame it more broadly - "visual analytics along specific use cases". 3.2
The general topic of analytics is a multidisciplinary area with methods, models and technologies from computer science, mathematics, economics, law, social sciences and more.
Education is the key to withstand the losses in jobs through automation in e.g. smart
manufacturing. Analysis helps to assess information simplified and provide suitable
information in time in the right proportion. Consequently, the mitigation in complexity
helps in education by accelerating training and further education [
Technological advances are able to transform public and societal services such as
health care [
Analysis comprises both technology impact assessment and regulatory impact
assessment, as they are a major concern of governmental administration. The first aims
at anticipating possible positive or negative impacts of an analytical result on society
[
Moreover, analytics may be applied in a single decision-making process to improve insights to the related problems and questions raised with the aim to improve transparency of such processes for decision-makers and stakeholders, as well, by improving understandability of the data. Moreover, given the fact that governmental decisions are flanked strongly with participation processes aiming to achieve acceptable and therefore better decisions, analytics may also play a key role in improving the participation process itself with regard to the stakeholder involvement, by improving the analytic capacities with regard to the given comments and their uptake, their relation to scientific or political documents, for both the decision makers and the public, as well.
Knowledge from management, such as innovation, technology and information management, but also strategic planning and predictions in companies (corporate foresight) should be considered. 3.3
Right on time is key in industry. Either if it is the right business decision in the business
strategy or if it is the opportunity to educate the professionals to strengthen their
personal journey. Analytics has impacts on both sides of industry, society and businesses.
Business and Society
8
Analytics helps to unveil formerly unknown technologies and their potential to deduct
better revised business decisions in terms of long planning business goals. Therefore,
help to build up a sustainable business strategy in shorter time is a huge advantage on
the market. While the development of enterprise strategies covers longer periods of
time (often month or years), there are a number of scenarios where analysis should help
to act within hours or month. The transportation and logistics domain have to react in
short times if a duty is cancelled or a partner resigns or cancels a contract. Only with
smart analytical technologies it is possible to react on those situations and find
alternative duties or immediately calculate cheaper or more efficient routes [
Furthermore, analytics helps train professionals to build faith in their companies as a source of motivation, which affects the company's economic side. So, effectiveness is another major aspect of economics. May it be from perspectives of highly skilled and professional employees or may it be with regard to retrieved insights in certain analysis. Only a high degree of professionalism and quality ensures decisions that support enterprises in being profitable. In perspective of data analytics this means that the entire processing pipeline has to be designed towards high data quality in perspective of completeness, cleansing and data interlinking, but also the visualization has to follow strict requirements in perspective of traceability, accuracy and clarity.
Lastly, the costs and earnings are essential criteria. To achieve these criteria, the impact of an analytical solution must lay on generating higher profits or lower the costs. From the research perspective the intention to increase the profits is often the smarter more creative way since this often comes along with building or entering new markets with new products. And this means, analytical solutions have to show, which markets this could be and what the major entering criteria will be. Since any market entering has a risk, even more analytical solutions are essential that are able to highlight risks (e.g. possible patent or legal problems in the entered market or country). In sum, there is wide range of scenarios where analytical solutions could help to lower the costs and/or increase incomes, and all of them could help companies to stay productive. 4
The organization of the planned analytics collaboration action consists of two main different phases (see Fig. 1).
The first is the proposal and negotiation phase, which also covers the team building that will work together on the specific goals. This first phase can be named as macrolevel perspective, since it majorly focuses on the definition and work out the basic goals and objectives.
The second phase is the concrete collaborative work, with the main goal to achieve new insights and technologies. Since the analytics collaboration action is majorly ICT driven, some outcomes will be in form of software prototypes. Due to the heterogeneity of proposed consortium, it is rather challenging to collaborate as far as no common understanding of use-case requirements are given. This perspective can also be named as micro-level perspective, since it aims to specify concrete use-cases and use-case requirements.
The figured process model (Fig. 1) consists about a hard-distinguished macro-level and micro-level perspectives, but it is to mention that in praxis it is usually not possible to split them that strict. Due to changes on project objectives and goals over time, e.g. due to partner changes, contextual changes, new research insights etc. some objectives at the macro-level can change with an impact as well as on the micro-level. So even when the micro-level phase already started, there could be changes on the macro-level as well. In sum, the process model in the presented form shows an ideal process model, which in praxis can regularly underlay some modifications. Furthermore, the process can be seen as an iterative process with the aim to refine analytics over many years in various iterations. 4.1
The macro-level perspective is particularly at the beginning of such as proposal phase and at the beginning of an accepted analytics collaboration action of relevance. The intention is to define common goals and objectives and teams to collaborate with to achieve these. At this level specific procedures, technologies, algorithms etc. are of interest and the focus lays only on the intended result and main outcome. This means also, that the knowledge/expertise and collaboration is important.
This principle intention is also considered in our defined analytics collaboration action application process, which is divided into four phases (see also Fig. 2).
Phase 1 aims at promoting discussions on Part B of the present document. The coordinator sends around the present document to inform the partners about the process and gives its proposition on the common understanding (Part B). The common understanding will be updated after comments, additions and requested changes of the proposal. Phase 1 concludes with an online-workshop to discuss and finalize the common understanding, to elect a management board, which will facilitate the drafting of the Business and Society 10 technical annex (Phase 2; see Part C) and to give further instructions on the application process (e.g. mode of further operation, answering questions etc.).
Phase 2 aims at drafting the technical annex (Part C), which forms the most sensitive part of the application. The drafting of the technical annex will be highly collaborative and the single members of the management board will be assigned specific tasks to fulfill. The draft will be finalized by the end of September, to allow for an assessment period (Phase 4).
Phase 3, which will be carried out in parallel to phase 2, aims at starting the analytics collaboration action application. This is considered to be the formal side of the application, which gives to the association to assess the conformity of the action and possible overlaps of the action with other actions on basis of a summary. Alterations are still possible in this phase. The coordinator will start the process shortly after and on the basis of the outcomes of phase 1. The partners of the analytics collaboration action will then be electronically invited to provide their contact and institutional information to the analytics collaboration action application. Phase 3 will iteratively inform phase 2 and vice versa.
Phase 4 aims at giving the partners enough time for a last internal assessment phase
before the final submission. Here, a generous time buffer is implemented, if one of the
other phases might require a longer period.
While the macro-level perspective is almost considering major goals and objectives and
to define principal aspects, the micro-level perspective is to achieve and a final
realization, especially in larger collaborating teams. Since in the work within heterogenous
teams do collaborate, different opinions and (technical) understandings are a common
challenge that we aim to phase with this micro-level process – also named as Use Case
Requirement Analysis Model (UCRA) [
Overall this model can be seen as an iterative approach. After the first round of the UCR analysis and implementation of discussed features, another round can be initiated to refine and extend the system.
One important fact is that this approach should be performed before major developments in perspective of user related features are in progress.
The paper aimed on describing how a strategic network should be founded that aims on optimizing and using technological and methodological innovations to identify analytical approaches in the areas of politics, business, education and research and to develop adequate solutions to the identified challenges. That strategic network of partners should rely on the expertise of the involved partners, who are internationally renowned in their respective fields. From a technological point of view, methods of digitization, artificial intelligence, visual analytics, data analytics and simulation should be considered to optimize and use for different application scenarios. Thereby, the actors have to respect the named challenges, but can precise and complete them further by themselves, whereby data-driven trends will also be identified. A following validation of the results is also carried out by the actors of the respective areas later on. Business and Society 12
This work was partially funded by the Federal Ministry of Education and Research under the grant number 13FH020AN9 and was conducted within the research group on Human-Computer Interaction and Visual Analytics (http://www.vis.h-da.de). For further information about the strategic alliance see https://www.vis.h-da.de/network.html.