Building and Exploring Dynamic Topic Models on the Web

Derntl, Michael; Günnemann, Nikou; Tillmann, Alexander; Klamma, Ralf; Jarke, Matthias

doi:10.1145/2661829.2661833

Cited by 6 publications

(3 citation statements)

References 5 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…While the basic LDA approach ignores word order, there are techniques that relax this constraint and consider limited forms of word‐order, for example, letting a word's presence depend on the preceding words . Other constraints get relaxed when applying LDA in dynamic scenarios such as finding time‐varying topics in a document stream . The topics themselves may be specified as being correlated, as in a hierarchy of topics .…”

Section: Topic Modelsmentioning

confidence: 99%

See 1 more Smart Citation

Emerging directions in predictive text mining

Indurkhya

2015

WIREs Data Min & Knowl

View full text Add to dashboard Cite

In recent years, Text Mining has seen a tremendous spurt of growth as data scientists focus their attention on analyzing unstructured data. The main drivers for this growth have been big data as well as complex applications where the information in the text is often combined with other kinds of information in building predictive models. These applications require highly efficient and scalable algorithms to meet the overall performance demands. In this context, six main directions are identified where research in text mining is heading: Deep Learning, Topic Models, Graphical Modeling, Summarization, Sentiment Analysis, Learning from Unlabeled Text. Each direction has its own motivations and goals. There is some overlap of concepts because of the common themes of text and prediction. The predictive models involved are typically ones that involve meta‐information or tags that could be added to the text. These tags can then be used in other text processing tasks such as information extraction. While the boundary between the fields of Text Mining and Natural Language Processing is becoming increasingly blurry, the importance of predictive models for various applications involving text means there is still substantial growth potential within the traditional sub‐fields of text mining. These data‐centric directions are also likely to influence future research in Natural Language Processing, especially in resource‐poor languages and in multilingual texts. WIREs Data Mining Knowl Discov 2015, 5:155–164. doi: 10.1002/widm.1154 This article is categorized under: Algorithmic Development > Text Mining Technologies > Prediction Technologies > Structure Discovery and Clustering

show abstract

Section: Topic Modelsmentioning

confidence: 99%

“…19 Other constraints get relaxed when applying LDA in dynamic scenarios such as finding time-varying topics in a document stream. 20 The topics themselves may be specified as being correlated, as in a hierarchy of topics. 21,22 These are all examples of novel scenarios that require enhancements to the basic LDA technique.…”

Section: Topic Modelsmentioning

confidence: 99%

Emerging directions in predictive text mining

Indurkhya

2015

WIREs Data Min & Knowl

View full text Add to dashboard Cite

show abstract

“…There are several exploratory tools that provide graphical user interfaces for end-users to explore topics in a collection of texts. While most of these tools use a static topic model such as LDA, Derntl et al (2014) provide a web-based application to explore topics using a dynamic topic model in reasonable runtime. In this research, two methods of LDA and DTM have been used to explore topics and their trends.…”

Section: Visualizing the Results Of Lda And Dtmmentioning

confidence: 99%

Approach to Identify Topics in a Collection of TIM Review Articles and their Changes Over Time

Nazari¹

View full text Add to dashboard Cite

Topic modeling can help better understand the content of large collections of text. The objective of the research is to develop an approach to identify latent topics in the Technology Innovation Management Review journal (TIM Review) and how topics have evolved using the LDA (Latent Dirichlet Allocation) and DTM (Dynamic Topic Model) algorithms. We applied the two approaches to a collection of TIM Review articles published between 2007 and 2017. We also examined extracted topics, most associated articles, topic labeling, topic/theme trends, and word trends produced by both approaches and identified the value of each approach. According to the results obtained, we identified 47 topics and categorized them into ten themes: open source, entrepreneurship, innovations, living labs, social technology innovation, growth, co-creation, cybersecurity, research, and ecosystem. While some topic trends became prominent over time, others disappeared. The distribution of the articles across topics in the LDA approach has been made more decisively so that of 597 articles, 503 most associated articles were identified, while this number is 299 articles in DTM. Furthermore, we discussed weaknesses and strengths of the algorithms to compare the performance of the two approaches based on defined criteria. We conclude that DTM provides more accurate word and topic trends over time, although it requires time slice settings and has a longer run-time compared to LDA. Finally, we document a set the steps of a process to carry out topic modeling analysis.

show abstract