Conversational Pattern Mining Using Motif Detection

AbstractThe subject of conversational mining has become great interest recently due to the explosion social and other online media. Supplementing this text is advancement in pre-trained language models which have helped us leverage these sources information. An interesting domain analyse conversations terms complexity value. Complexity arises fact that a conversation can be asynchronous involve multiple parties. It also computationally intensive process. We use unsupervised methods our work order develop pattern technique does not require time consuming, knowledge demanding resource labelling exercises. The task identifying repeating patterns sequences well researched Bioinformatics field. In work, we adapt field Natural Language Processing make several extensions motif detection algorithm. demonstrate application algorithm on dynamic, real world data set; extract motifs from an open-source film script source. run exploratory investigation into types are able mine.KeywordsNatural ProcessingPattern miningMotif miningConversational analysis