Andere Kunden interessierten sich auch für
![Learning Users' Profiles in Location-based Social Networks]( "Learning Users' Profiles in Location-based Social Networks")
Learning Users' Profiles in Location-based Social Networks50,99 €![Biometrics for User Authentication Using Artificial Neural Networks]( "Biometrics for User Authentication Using Artificial Neural Networks")
Biometrics for User Authentication Using Artificial Neural Networks36,99 €![Routing in Heterogeneous Wireless Sensor Networks]( "Routing in Heterogeneous Wireless Sensor Networks")
Routing in Heterogeneous Wireless Sensor Networks23,99 €![Extending the FTT-SE Protocol for Multi-Master/Multi-Slave Networks]( "Extending the FTT-SE Protocol for Multi-Master/Multi-Slave Networks")
Extending the FTT-SE Protocol for Multi-Master/Multi-Slave Networks32,99 €![Characteristics of Neural Networks based on Energy Functions]( "Characteristics of Neural Networks based on Energy Functions")
Characteristics of Neural Networks based on Energy Functions36,99 €![Role of ANN Techniques in Detection of Breast Cancer]( "Role of ANN Techniques in Detection of Breast Cancer")
Role of ANN Techniques in Detection of Breast Cancer32,99 €![Stochastic Weight Update in Neural Networks]( "Stochastic Weight Update in Neural Networks")
Stochastic Weight Update in Neural Networks32,99 €
Sentiment analysis is a new field of research that is getting very popular due to the demand to understand what are people's thoughts or opinions in the internet. There are major challenges to understand natural language and how to classify opinions with high accuracy. To classify student opinions a system has been created that approaches the problem in two ways, using fixed-rules and using machine learning algorithms. The classifier that are created are able to classify student opinions as positive and negative. The dataset that created contains more than 250 student opinions that were gathered and processed. Evaluation of various algorithms that were used to classify opinions is performed in order to learn what will suit best the case with the current data-set. The system uses excessively dictionaries with positive and negative words when fixed-rules approach is used, and the data-set's are used for other machine learning algorithms. Input is processed in lexical and syntactical level before it is used to train the model and to be classified. The system that we chose to receive student opinions and to use the classifiers is Effectinet, developed by a student at CITY College.