{"title":"初步注解结果:双边对话的议价角色","authors":"Md. Alvee Islam Navid, K. Fujita","doi":"10.1109/IIAI-AAI50415.2020.00115","DOIUrl":null,"url":null,"abstract":"Negotiations are common; however, mining important aspects from such dialogues is challenging because the lack of a proper strategy and the use of predetermined words, as well as sentences, could lead to on-going disputes or failure. In addition, existing studies have focused on bargaining flows but have neglected the semantic and negotiation roles of negotiations. As a first step in negotiation mining, this paper proposes an annotation scheme to obtain the maximum number of labeled segments in negotiation sessions. We focus on the unique negotiation roles used to come to agreements in bilateral bargaining. We define five labels in the negotiations: Greet, Inquiry, Statement, DialogAct, and Other. These labels include attributes and features that consider the roles of the labels. We conducted the annotation by recruiting four annotators, who performed the annotations after reading the guidelines and attending several training meetings. After the dataset was divided into 876 dialogues, each dialogue was annotated by two annotators. We evaluate the Cohen kappa measures of the agreement between two annotations and achieve a sufficient score to analyze the negotiation roles in the dialogues. In particular, DialogAct, which is important to detect negotiation actions, achieved a high score with respect to understanding the flow of the negotiations.","PeriodicalId":188870,"journal":{"name":"2020 9th International Congress on Advanced Applied Informatics (IIAI-AAI)","volume":"20 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2020-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Preliminary Annotation Results: Bargaining Roles for Bilateral Dialogues\",\"authors\":\"Md. Alvee Islam Navid, K. Fujita\",\"doi\":\"10.1109/IIAI-AAI50415.2020.00115\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Negotiations are common; however, mining important aspects from such dialogues is challenging because the lack of a proper strategy and the use of predetermined words, as well as sentences, could lead to on-going disputes or failure. In addition, existing studies have focused on bargaining flows but have neglected the semantic and negotiation roles of negotiations. As a first step in negotiation mining, this paper proposes an annotation scheme to obtain the maximum number of labeled segments in negotiation sessions. We focus on the unique negotiation roles used to come to agreements in bilateral bargaining. We define five labels in the negotiations: Greet, Inquiry, Statement, DialogAct, and Other. These labels include attributes and features that consider the roles of the labels. We conducted the annotation by recruiting four annotators, who performed the annotations after reading the guidelines and attending several training meetings. After the dataset was divided into 876 dialogues, each dialogue was annotated by two annotators. We evaluate the Cohen kappa measures of the agreement between two annotations and achieve a sufficient score to analyze the negotiation roles in the dialogues. In particular, DialogAct, which is important to detect negotiation actions, achieved a high score with respect to understanding the flow of the negotiations.\",\"PeriodicalId\":188870,\"journal\":{\"name\":\"2020 9th International Congress on Advanced Applied Informatics (IIAI-AAI)\",\"volume\":\"20 1\",\"pages\":\"0\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2020-09-01\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"2020 9th International Congress on Advanced Applied Informatics (IIAI-AAI)\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/10.1109/IIAI-AAI50415.2020.00115\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"2020 9th International Congress on Advanced Applied Informatics (IIAI-AAI)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/IIAI-AAI50415.2020.00115","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
Preliminary Annotation Results: Bargaining Roles for Bilateral Dialogues
Negotiations are common; however, mining important aspects from such dialogues is challenging because the lack of a proper strategy and the use of predetermined words, as well as sentences, could lead to on-going disputes or failure. In addition, existing studies have focused on bargaining flows but have neglected the semantic and negotiation roles of negotiations. As a first step in negotiation mining, this paper proposes an annotation scheme to obtain the maximum number of labeled segments in negotiation sessions. We focus on the unique negotiation roles used to come to agreements in bilateral bargaining. We define five labels in the negotiations: Greet, Inquiry, Statement, DialogAct, and Other. These labels include attributes and features that consider the roles of the labels. We conducted the annotation by recruiting four annotators, who performed the annotations after reading the guidelines and attending several training meetings. After the dataset was divided into 876 dialogues, each dialogue was annotated by two annotators. We evaluate the Cohen kappa measures of the agreement between two annotations and achieve a sufficient score to analyze the negotiation roles in the dialogues. In particular, DialogAct, which is important to detect negotiation actions, achieved a high score with respect to understanding the flow of the negotiations.