ConVis: A Visual Text Analytic System for Exploring Blog Conversations

Abstract

Today it is quite common for people to exchange hundreds of comments in online conversations (e.g., blogs). Often, it can be very difficult to analyze and gain insights from such long conversations. To address this problem, we present a visual text analytic system that tightly integrates interactive visualization with novel text mining and summarization techniques to fulfill information needs of users in exploring conversations. At first, we perform a user requirement analysis for the domain of blog conversations to derive a set of design principles. Following these principles, we present an interface that visualizes a combination of various metadata and textual analysis results, supporting the user to interactively explore the blog conversations. We conclude with an informal user evaluation, which provides anecdotal evidence about the effectiveness of our system and directions for further design.

Collective Sensemaking in Online Health Forums

“…In computational linguistics, new methods focus on the design of novel statistical models for tasks such as event detection [2], sentiment analysis [29], and modeling of the topics of discussions [33] as well as predicting the structure of discourse [39]. Taking these trends further, Hoque et al used a combination of NLP and data visualization techniques to visualize salient properties of discussion threads, such as topics discussed and sentiment towards these topics [15]. The framework suggested in this paper can provide foundation for new approaches to facilitating collective sensemaking using Understanding Health through Online Behavior CHI 2015, Crossings, Seoul, Korea computational text analysis and interactive visualizations.…”

VisOHC: Designing Visual Analytics for Online Health Communities

“…For example, ConVis [17] visualized a single thread of conversation, instead of the whole online community. By doing this, a user can follow how conversation evolves among participating members, which could provide a much richer context as compared to showing only some key phrases.…”

“…Since the design illustrates individual threads in a horizontal slice of a viewport, it is most effective for exploration and detailed analyses, but is not effective for overview tasks. If our domain tasks follow the traditional visual information seeking mantra, we might as well visualize an overview of the entire thread, such as, in terms of topics, e.g., in [17]. In fact, at an initial stage, we also considered a ThemeRiver-style temporal trend visualization [15] using topic modeling [6], which should provide a topical summary of the entire thread.…”

“…Prior studies investigated the use of visual analytic tools to explore OHC user postings in terms of temporal trends and sentiments [7] and relationships between members and postings through network analysis [16]. Studies on online communities also provided insights that facilitate an understanding of OHC conversations, through means such as topic analyses of discussion threads [25] and authorship networks on top of topics [17]. Although these approaches are inspiring, their design is not intended for OHC administrators, who face adverse events, such as suicidal and emotional messages, as well as patients requiring urgent help.…”

Interactive Exploration of Asynchronous Conversations: Applying a User-centered Approach to Design a Visual Text Analytic System

“…We first characterize the domain of blogs and perform the data and tasks abstraction according to the nested model of design study (Munzner, 2009). We then mine the data as appeared to be essential from that data and task analysis, followed by iteratively refining the design of ConVis that aims to effectively support the identified blog reading tasks (A more detailed analysis of the task abstractions and visual design is provided in (Hoque and Carenini, 2014)). …”