Browse Prior Art Database

System for commmunicator BUSY status automatic prediction Disclosure Number: IPCOM000246412D
Publication Date: 2016-Jun-06
Document File: 2 page(s) / 45K

Publishing Venue

The Prior Art Database


When you are extremely in hurry; trying to finish urgent tasks, seeking for minute of peace you online communicator is interrupting you. The core of the proposal is adoption of machine learning algorithms to predict busy state based on historical data.

This text was extracted from a PDF file.
This is the abbreviated version, containing approximately 100% of the total text.

Page 01 of 2

System for commmunicator BUSY status automatic prediction Busy state prediction can be done as follows:

1. Historical data used as training data

                training data . In historical data classification label : [BUSY, NOT BUSY] is gathered either manually by the user (system asks for confirmation) or automatically based on manual switch to busy status on communicator. The list list of the monitored attributes (feature set) is described in table 1.

Figure 1. Flow chart of system

Table 1. TRAINING DATA - List of features and sample values


Page 02 of 2

2. TRAINING DATA is used to train model. Machine learning algorithms can be used; for example: Support Vector Machines. As a result of training model is created.

3. CLASSIFICATION DATA - data that is currently gathered (feature set) and [BUSY, NOT BUSY] status is unknown. Previously trained model is applied and prediction is made

4. Based on prediction/classification result communicator status is either changed to [BUSY] or [NOT BUSY]

5. SELF-CORRECTNESS - system provides option for user feedback (in case of incorrect prediction). In such case such update will extend the training set and improve classification correctness in future .