Bianka Trevisan

Part-Of-Speech Tagging for Social Media Texts

Work on Part-of-Speech (POS) tagging has mainly concentrated on standardized texts for many years. However, the interest in automatic evaluation of social media texts is growing considerably. As the nature of social media texts is clearly different from standardized texts, Natural Language Processing methods need to be adapted for reliable processing. The basis for such an adaption is a reliably tagged social media text training corpus. In this paper, we introduce a new social media text corpus and evaluate different state-of-the-art POS taggers that are retrained on that corpus. In particular, the applicability of a tagger trained on a specific social media text type to other types, such as chat messages or blog comments, is studied. We show that retraining the taggers on in-domain training data increases the tagging accuracies by more than five percentage points.

New ways to develop professional communication concepts

This paper presents an interdisciplinary approach concerning a professional communication concept for energy projects using the example of deep geothermal energy. The interdisciplinary approach combines methods and approaches of engineering science, psychology, and professional communication. The communication concept is based on empirical case studies. It addresses decision makers (e.g. project leaders or mayors), communication professionals working for energy projects, and interested citizens. It aims to enable project managers and communication experts to establish an open-minded dialog between both sides (project leader and involved stakeholders) that fosters the public acceptance of renewable energy projects and helps interested people to inform themselves about the given technology. The communication strategy consists of a communication concept, background information, guidelines, and a Web app. The strategy supports information and communication tasks along the whole project life cycle and reacts on site- or event-specific needs. The communication strategy can be adapted to support the introduction and operation of other technologies.

Focused crawling for building Web comment corpora

Web 2.0 provides various types of social media applications, e.g., blogs, forums and news sites that allow users to post Web comments. This kind of communication plays an important role in acceptance research. To extract different opinions from such data, it is necessary to build Web comment corpora. Building such corpora requires focused crawling. Many focused Web crawling algorithms are known to build topic-specific Web collections. However, the type of Web pages is typically not considered. In this paper, we introduce a new type-specific focused crawler, which uses a classifier based on HTML meta information. Its application allows for collecting only Web pages that cover Web comments from various domains.

Talking about mobile communication systems: verbal comments in the web as a source for acceptance research in large-scale technologies

In the research project HUMIC (RWTH Aachen University) we strike out in a new direction in acceptance research. In order to identify previously undetected acceptance factors in user-generated content, traditionally used methods are complemented by innovative methods of computational linguistics. Verbal comments from social media tools like weblogs are analyzed by text mining methods with the aim to get access to user judgements. This methodology offers the possibility to learn how user perceive and conceptualize large-scale technologies using the example of the mobile communication systems.

Web comment-based trend analysis on deep geothermal energy

In this paper we present the initial results of a national trend analysis - an approach that allows collecting and investigating location- and time-specific acceptance factors from user-generated content. For this purpose, an annotation scheme is adapted that is originally developed for sentiment analysis and opinion detection. By applying this annotation scheme, German Web comments of a newspaper and a news-site are quantitatively and qualitatively analyzed. The analysis focuses on the investigation of acceptance drivers of deep geothermal energy. Thereby, it is assumed that the public opinion - positive, negative or neutral - is location-and time-dependent. In contradiction, media often draw an adulterated picture of citizen opinions. Our initial assumption of opposed opinions on deep geothermal energy in public and media was confirmed by the conducted trend analysis.

Medusa and Pandora meet the web 2.0: How risk types influence the communication in social media

Risk communication is an important branch of technical communication: Innovative technologies need to be communicated to the general public; concerns of relevant stakeholders need to be addressed. Such concerns can be identified in topic-related comments in social media. This paper presents the results of a study on how people evaluate technologies that are characterized by as different risk types. Two examples of technology-induced risks types are examined: (1) The ubiquitous and persistent impacts of hydraulic fracturing (risk type: Pandora). (2) The consequences with high exposure but little damage potential of a local infrastructure project aiming at the introduction of a light-rail system (risk type: Medusa). Risk-relevant comments were identified automatically using a topic tracking approach based on dictionaries of risk-related terms. The identified comments were analyzed quantitatively and qualitatively with regard to comment types, formulation styles, evaluation types, evaluated aspects and objects of comparison. The results show that the type of risk influences which aspects of a technology users evaluate and how they formulate evaluations. The findings are discussed with respect to implications for risk communication.

Indicators of Argument-conclusion Relationships. An Approach for Argumentation Mining in German Discourses

[In this paper we present a new methodological approach for the analysis of public discourses aiming at the semi-automated identification of arguments by combining methods from discourse analysis with methods from Natural Language Processing. Discourses evolve over long periods of time and, consequently, form a broad database. Up to now, the analysis of discourses is hitherto performed primarily by hand, i.e., only small corpora or discourse fragments can be analyzed. Inevitably, this leads to lengthy and expensive annotation. Thus, there is a growing interest to overcome these methodological challenges by the use of computer-based methods and tools for the semi-automated analysis. However, there are only few approaches known that focus on the analysis of discourses and the (semi-)automated identification of arguments therein (e.g. Reed at al., 2008; Liakata et al., 2012; Ashley and Walker, 2013). Particularly, approaches that can be explicitly used for the analysis of German-language discourses exist only in initial stages. Therefore, we suggest a fine-grained semi-automated approach based on multi-level annotation that focuses on linguistic means as indicators of arguments. The aim is to identify regularities, respectively, indicators in the linguistic surface of the discourse (e.g. recurring lexical and typographical characteristics), which indicate the occurrence of certain arguments (e.g. premise). In this paper, we focus on the identification of indicators of argumentconclusion relationship: conclusive connectors or conclusiva, that are typically adverbs such as hence, consequently, therefore, thus, because (Govier, 2013; see example below): Die Campusbahn werde den Individualverkehr verdrangen, weil die Stadt eng bebaut sei. Schon in den 1970er Jahren sei deshalb das Aus fur die Strasenbahn besiegelt worden. [The campus train will displace the individual traffic because the city is densely built. Therefore, the end for the tram was sealed in the 1970s.] As an application example, a small corpus consisting of 21 newspaper articles is analyzed. The corpus belongs to the interdisciplinary project Future Mobility (FuMob), which is funded by the Excellent Initiative of the German federal and state governments. The methodological approach consists of three steps, which are performed iteratively: (1) manual discourselinguistic argumentation analysis, (2) semiautomatic Text Mining (PoS-tagging and linguistic multi-level annotation), and (3) data merge. (1) Discourse-linguistic argumentation analysis: First, the data is manually analyzed. Objectives of the analysis are (i) identifying discourserelevant arguments, (ii) forming argument classes, and (iii) determining the significance of an argument in the discourse (Niehr 2004). To determine the significance of an argument the use by various discourse participants is analyzed and quantified. The argument-use can be categorized as argumentative, positively cited, negatively cited or neutrally cited. In addition, to identify arguments and their use in public discourse, the analysis also aims to detect and characterize discourse participants who use similar arguments.

Communication of new energy forms: Ways to detect topics and stakeholders

This paper addresses professional communication experts who develop communication concepts for companies and the introduction of new technologies, e.g., in the energy sector. It presents an approach exploiting social media applications (Facebook®, blogs) as an information resource for the communication of technologies by detecting information on relevant topics and stakeholders. For the topic and stakeholder detection, text mining-methods are applied. As an application example, new energy forms serve. The results show exemplarily that social media applications, such as blogs provide a wealth of information on relevant topics in public discourse, while social networking sites such as Facebook® are particularly useful for the detection of stakeholder profiles. The proposed approach can help communication experts to retrieve and gain information, for instance, as input for communication concepts and strategies.

Eliciting user requirements and acceptance for customizing mobile device system architecture

Mass customization is a popular approach in product design and manufacturing, where customers can configure standard products according to their individual preferences. Applied to the technical customization of mobile device system architecture (e.g. smartphones), an empirical multi-method approach was applied in order to elicit user requirements and acceptance. First, in a text mining analysis with n=80.995 blog comments relevant components and properties of cell phones were identified. Second, an online-survey with n=48 participants was conducted, which quantified user requirements and acceptance of the customization approach. The consecutive combination of text mining and survey provided valuable insights into user perceptions and acceptance. Customization was perceived positively, although the willingness to pay was low. Customizable technical characteristics in mobile device system design such as battery life, speech quality, memory capacity and connection quality as well as user profiles were identified.

Evaluation of user-centred methods in product design: a holistic approach

The user’s preference on a product depends on various factors such as the hedonic and pragmatic product performance as well as individual requirements. Thus, to ensure user-centred product design, methods must be identified which support the designer in taking the user perspective into account and promote the development of a holistic design integrating pragmatic and hedonic aspects. In the project ‘Gender-specific Kansei Engineering’ five methods for the identification of user-centred product concepts and quality requirements are tested in three target groups (users, experts, and designers). The methods are evaluated on two dimensions: data quality and application effort. The evaluation shows that qualitative methods (couple interview and focus group) are in terms of data quality more helpful in product design. In contrast, quantitative methods (AttrakDiff2, Kano-questionnaire, and Prototype mapping) generate lower application effort.