Elements of information theory
Elements of information theory
Improved boosting algorithms using confidence-rated predictions
COLT' 98 Proceedings of the eleventh annual conference on Computational learning theory
Large Margin Classification Using the Perceptron Algorithm
Machine Learning - The Eleventh Annual Conference on computational Learning Theory
Working Set Selection Using Second Order Information for Training Support Vector Machines
The Journal of Machine Learning Research
Relational learning for email task management
IJCAI'05 Proceedings of the 19th international joint conference on Artificial intelligence
LIBSVM: A library for support vector machines
ACM Transactions on Intelligent Systems and Technology (TIST)
Hi-index | 0.00 |
Diverse social networks and online communities add more messages to catch up to active communication users who are already flooded by emails, SMS, and IMs. We present application frameworks and methods that aim to assist the users to efficiently identify important messages from diverse interactive message channels, in particular, as an example, the messages containing actionable items. To build the method for identifying the messages containing actionable items, an empirical study on choosing the machine learning based classification algorithm and construction of the best feature set was conducted, illustrating the various aspects to consider for the best classification performance. A set of novel and elegant rules for parsing and extracting a summary of the actionable items from the identified message are presented and its effectiveness is demonstrated through examples and measurements on an email data set.