Modeling the Chatbot Quality of Services (CQoS) Using Word Embedding to Intelligently Detect Inappropriate Responses
- Publisher:
- Springer
- Publication Type:
- Conference Proceeding
- Citation:
- Advanced Information Networking and Applications, 2020, 1151, pp. 60-70
- Issue Date:
- 2020-01-01
Closed Access
Filename | Description | Size | |||
---|---|---|---|---|---|
Almansor-Hussain2020_Chapter_ModelingTheChatbotQualityOfSer.pdf | 1.76 MB |
Copyright Clearance Process
- Recently Added
- In Progress
- Closed Access
This item is closed access and not available.
The rapid growth of intelligent chatbots as conversational agents with the assistance of artificial intelligence has recently attracted much research attention. The major role of a chatbot is to generate appropriate responses to the user, however sometimes the chatbot fails to understand the user’s meaning. Therefore, detecting inappropriate responses from a chatbot is a critical issue. Several studies based on annotated datasets have investigated the problem of inappropriate responses from a chatbots perspective without considering the user’s perspective. Understanding the context of the conversation is an important point in determining whether a response is appropriate or inappropriate. Sentiment analysis is a natural language processing task that supports mining in user behavior. Therefore, we propose an intelligent framework that combines automated sentiment scoring and a word embedding model to detect the quality of chatbot responses considering the end-user’s point of view. We find our model achieves higher quality results than logistic regression.
Please use this identifier to cite or link to this item: