Tandem Anchoring: a Multiword Anchor Approach for Interactive Topic Modeling

Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics, pages 896–905 Vancouver, Canada, July 30 - August 4, 2017. c⃝2017 Association for Computational Linguistics https://doi.org/10.18653/v1/P17-1083 Proceedings of the 55th Annual Meeting of the Association for Computational Linguistics, pages 896–905 Vancouver, Canada, July 30 - August 4, 2017. c⃝2017 Association for Computational Linguistics https://doi.org/10.18653/v1/P17-1083 Tandem Anchoring: a Multiword Anchor Approach for Interactive Topic Modeling Jeffrey Lund, Connor Cook, Kevin Seppi Computer Science Department Brigham Young University {jefflund,cojoco,kseppi}@byu.edu Jordan Boyd-Graber Computer Science Department University of Colorado Boulder jordan.boyd.graber@colorado.edu Abstract Interactive topic models are powerful tools for understanding large collections of text. However, existing sampling-based inter- active topic modeling approaches scale poorly to large data sets. Anchor meth- ods, which use a single word to uniquely identify a topic, offer the speed needed for interactive work but lack both a mecha- nism to inject prior knowledge and lack the intuitive semantics needed for user- facing applications. We propose combina- tions of words as anchors, going beyond existing single word anchor algorithms— an approach we call “Tandem Anchors”. We begin with a synthetic investigation of this approach then apply the approach to interactive topic modeling in a user study and compare it to interactive and non- interactive approaches. Tandem anchors are faster and more intuitive than existing interactive approaches. Topic models distill large collections of text into topics, giving a high-level summary of the the- matic structure of the data without manual anno- tation. In addition to facilitating discovery of top- ical trends (Gardner et al., 2010), topic modeling is used for a wide variety of problems including document classification (Rubin et al., 2012), in- formation retrieval (Wei and Croft, 2006), author identification (Rosen-Zvi et al., 2004), and senti- ment analysis (Titov and McDonald, 2008). How- ever, the most compelling use of topic models is to help users understand large datasets (Chuang et al., 2012). Interactive topic modeling (Hu et al., 2014) al- lows non-experts to refine automatically generated topics, making topic models less of a “take it or leave it” proposition. Including humans input dur- ing training improves the quality of the model and allows users to guide topics in a specific way, cus- tom tailoring the model for a specific downstream task or analysis. The downside is that interactive topic model- ing is slow—algorithms typically scale with the size of the corpus—and requires non-intuitive in- formation from the user in the form of must-link and cannot-link constraints (Andrzejewski et al., 2009). We address these shortcomings of interac- tive topic modeling by using an interactive version of the anchor words algorithm for topic models. The anchor algorithm (Arora et al., 2013) is an alternative topic modeling algorithm which scales with the number of unique word types in the data rather than the number of documents or tokens (Section 1). This makes the anchor algorithm fast enough for interactive use, even in web-scale doc- ument collections. A drawback of the anchor method is that anchor words—words that have high probability of …

Content truncated to 500 words