英语翻译Mixed-initiative in the thermostatIn the intelligent the

英语翻译
Mixed-initiative in the thermostat
In the intelligent thermostat a statistical model approach is
combined with a dialog manager and embedded task
model.The statistical model is used to determine the
likelihood of a trend in temperature settings.The certainty
of a given trend or pattern influence the dialog presented
to the user.A pattern is derived from:time and date,
room,frequency of presence in room,user preference for
temperature,and actual room temperature.With the
exception of desired room temperature all of the variables
can be directly determined by the system.Desired
temperature is inferred by considering the history of actual
room temperature settings,actual presence in a room and
whether or not a local heating or cooling control has been
adjusted.For example,if the bathroom heater is switched
on while the room is only occupied two hours a day,the
desired temperature for those two hours is based on the
temperature during those two hours.The system can
suggest a desired temperature setting and for how long the
bathroom heater should be switched on.The user can
directly increase or decrease the proposed temperature
setting and duration through spoken dialog with the
system or by direct touch-based manipulation of the
interface.
The system is currently being developed to interpret user
intentions and provide suggestions through a dialog
manager that recognises a number of likely tasks.
Currently,Collagen [4] is being considered as an
application-independent collaboration manager.The
collaboration manager will keep track of the linguistic and
attentional state of discourse and the collaborative
intentions of the participants (thermostat agent and user).
An embedded task model which contains recipes will be
developed in support of anticipated user goals.Recipes
specify the steps and constraints required to perform
thermostat control tasks.The recipes along with arbitrary
pattern-action rules are intended to serve as input to the
dialog manager

温度调节装置的组合自动界面
在智能化温度调节装置里头,一个数据模型的处理附带这对话管理跟嵌入式任务模式.这个数据模型是用来温度设定里头的可能趋势,趋势或模式的确核结果能够直接反映给用户,其模式分为：时间、数据（信息）、空间容量、空间显示频率,用户选择温度、实际空间温度,在预设空间温度值之外的,所有的变相参数都能够由系统确定.比方说,如果一个房间每天只用两个小时,浴霸（浴室加热装置）就会自动打开,温度预设值就会根据那两个小时内的环境温度来自动设定,预设温度值是根据过去的温度设定值、实际显示（产生）值、或有无加热和冷却控制调节来演推而得的,譬如,系统能够自动设定一个温度值,决定加热装置需要开启多久,用户能够通过系统对话操作跟直接点击应用界面来升高或降低希望的温度值.