Learning to Disentangle Interleaved Conversational Threads with a Siamese Hierarchical Network and Similarity Ranking

Proceedings of NAACL-HLT 2018, pages 1812–1822 New Orleans, Louisiana, June 1 - 6, 2018. c⃝2018 Association for Computational Linguistics Learning to Disentangle Interleaved Conversational Threads with a Siamese Hierarchical Network and Similarity Ranking Jyun-Yu Jiang†, Francine Chen‡, Yan-Ying Chen‡ and Wei Wang† †University of California, Los Angeles, CA, USA ‡FX Palo Alto Laboratory, Palo Alto, CA, USA jyunyu@cs.ucla.edu, {chen,yanying}@fxpal.com, weiwang@cs.ucla.edu Abstract An enormous amount of conversation occurs online every day, such as on chat platforms where multiple conversations may take place concurrently. Interleaved conversations lead to difficulties in not only following discussions but also retrieving relevant information from simultaneous messages. Conversation disen- tanglement aims to separate intermingled mes- sages into detached conversations. In this paper, we propose to leverage represen- tation learning for conversation disentangle- ment. A Siamese hierarchical convolutional neural network (SHCNN), which integrates local and more global representations of a message, is first presented to estimate the conversation-level similarity between closely posted messages. With the estimated simi- larity scores, our algorithm for conversation identification by similarity ranking (CISIR) then derives conversations based on high- confidence message pairs and pairwise redun- dancy. Experiments were conducted with four publicly available datasets of conversations from Reddit and IRC channels. The experi- mental results show that our approach signif- icantly outperforms comparative baselines in both pairwise similarity estimation and con- versation disentanglement. 1 Introduction With the growth of ubiquitous internet and mobile devices, people now commonly communicate in the virtual world. Among the various methods of communication, text-based conversational media, such as internet relay chat (IRC) (Werry, 1996) and Facebook Messenger1, has been and remains one of the most popular choices. In addition, many enterprises have started to use conversational chat platforms such as Slack2 to enhance team col- laboration. However, multiple conversations may 1Facebook Messenger: https://www.messenger. com/ 2Slack: https://slack.com/ Thread Message ... ... T31 Malcolm: If running as root, I need to set up a global config rather than ⇠/.fetchmailrc ? T38 Elma: i’m sure i missed something but fonts rendering in my gimp works isn’t at its best T39 Sena: is there anyway to see what the CPU temperature is? T38 Elma: is it because of gimp or i missed some tuning or such? T31 Rache: Specify a non-default name run control file. T41 Denny: so how does one enforce a permission set and own- ership set on a folder and all its children? T31 Malcolm: in the man page it doesn’t mention any global fetchmailrc file... that is what was confusing me... T42 Shenna: hi, are sata drives accessed as sda or hda? T41 Elma: -R for recursive... T42 Elma: sda ... ... Figure 1: A segment of real-world conversations in- volving six users and five (annotated) threads from the IRC dataset. occur simultaneously when conversations involve three or more participants. Aoki et al. (2006) found an average of 1.79 conversations among eight participants at a time. Moreover, some plat- forms like chatrooms in Twitch may have more concurrent conversations (Hamilton et al., 2014). Interleaved conversations can lead to …

Content truncated to 500 words