Robot Application In The Medical Area
Zhou, Yi (2019)
Julkaisun pysyvä osoite on
https://urn.fi/URN:NBN:fi:amk-2019052611935
https://urn.fi/URN:NBN:fi:amk-2019052611935
Tiivistelmä
With the development of Big Data and Artificial Intelligence, the robot is currently deployed in many areas. This thesis aims to develop a health assistant program running in Pepper robot. After the training process, the robot can answer some basic health questions.
This thesis is mainly divided into five parts: Technology background, Data collection, Data training, Response, and Result. Technology background indicates relative technical specification such as a brief introduction to Machine Learning and Chatterbot library. Data collection mainly introduce how to collect data and re-organize data. The training process is presented in the Data training part, all the data mentioned above is trained by the Chatterbot library training method and additional algorithms. The response process, which is one of the most important parts of this thesis, indicates how to get an answer based on the question the user asks. Result part shows a demo video and improvement point of this thesis.
Before deploying this project, the data should be prepared. The medical materials and conversation will be fetched from the health website. And then, all the data will be listed systematically and making them fitted for training. After that, the robot will be trained by the prepared data and algorithm. An intelligent robot will be presented when finishing the training. Through the “ALAudioRecord” module, the Pepper robot records the questions the user asks and then store them in the local space. Google Speech API will upload the voice file. Google Cloud analyzes the voice file and converts it into a text message. The algorithm splits messages and mark them. Finally, the Pepper robot gets a suitable answer based on the question.
Python is the programming language for this project, the version is Python 2.7 for Linux. Because of the limited version of the Naoqi system, all the code run under Ubuntu 16.
This thesis is mainly divided into five parts: Technology background, Data collection, Data training, Response, and Result. Technology background indicates relative technical specification such as a brief introduction to Machine Learning and Chatterbot library. Data collection mainly introduce how to collect data and re-organize data. The training process is presented in the Data training part, all the data mentioned above is trained by the Chatterbot library training method and additional algorithms. The response process, which is one of the most important parts of this thesis, indicates how to get an answer based on the question the user asks. Result part shows a demo video and improvement point of this thesis.
Before deploying this project, the data should be prepared. The medical materials and conversation will be fetched from the health website. And then, all the data will be listed systematically and making them fitted for training. After that, the robot will be trained by the prepared data and algorithm. An intelligent robot will be presented when finishing the training. Through the “ALAudioRecord” module, the Pepper robot records the questions the user asks and then store them in the local space. Google Speech API will upload the voice file. Google Cloud analyzes the voice file and converts it into a text message. The algorithm splits messages and mark them. Finally, the Pepper robot gets a suitable answer based on the question.
Python is the programming language for this project, the version is Python 2.7 for Linux. Because of the limited version of the Naoqi system, all the code run under Ubuntu 16.