International Conference on Advances in Computing and Technology (ICACT)
Permanent URI for this communityhttp://repository.kln.ac.lk/handle/123456789/15607
Browse
3 results
Search Results
Item Forecasting Monthly Ad Revenue from Blogs using Machine Learning(3rd International Conference on Advances in Computing and Technology (ICACT ‒ 2018), Faculty of Computing and Technology, University of Kelaniya, Sri Lanka., 2018) Dias, D.S.; Dias, N.G.J.Blogs emerged in the late 1990s as a technology that allows Internet users to share information. Since then, blogging has evolved to become a source of living to some and a hobby to others. A blog with rich content and regular traffic could easily be monetized through a number of methods. Affiliate marketing, Google AdSense, offering courses or services, selling eBooks and paid banner advertisements are some of the methods in which a blog could be monetized. There exists, a direct relationship on the revenue that can be generated through any of the above methods and the traffic that the blog gets. Google AdSense is the leader in providing ads from publishers to website owners. All bloggers or blogging website owners who have monetized their blogs, attempt to maximize their revenue by publishing articles in hope that it will generate the targeted revenue. On the other hand, bloggers or blogging website owners that hope to monetize their blog will be greatly benefitted if there was a way to forecast the monthly ad revenue that could be generated through the blog. But there exists no tool in the market that can help the bloggers forecast their ad revenue from the blog. In this research, we are looking at the possibility of finding an appropriate machine learning technique by comparing a linear regression, neural network regression and decision forest regression approaches in order to forecast the monthly ad revenue that a blog can generate to a greater accuracy, using statistics from Google Analytics and Google AdSense. As conclusion, the Decision Forest Regression model came out as the best fit with an accuracy of over 70%Item Virtual Airplay Drum Kit based on Hand Gesture Recognition(3rd International Conference on Advances in Computing and Technology (ICACT ‒ 2018), Faculty of Computing and Technology, University of Kelaniya, Sri Lanka., 2018) Dias, D.S.; Perera, M.D.R.In the music industry, a drum kit plays a vital role in the production of masterpiece musical melodies. It is also one of the instruments that is in greatest demand by youngsters who are passionate in learning and practicing music. But acquiring a typical drum kit is become a difficult task because of its high cost as well as it requires a large storage space to hold. This research is targeted in examining the possibility of engineering a cost-effective solution to build a portable drum kit. In this approach, ultrasonic sensors are used in order to identify hand gestures. Ultrasonic sensor is used to measure the distance to an obstacle using the theories of sound reflectance. The obstacle in this scenario is the human palm. When the palm of the human is moved up and down above the ultrasonic sensor, mimicking the typical actions of playing a drum kit, the changes in distances to the palm are mapped to corresponding drum sounds using a sound generation algorithm. This algorithm is further optimized in such a way that it yields an optimal consistency in readings, regardless of the typical issues of the low cost ultrasonic sensor such as noise, low accuracy of distance readings and random loss of signal. The solution was tested with the feedback of the general audience and it yielded satisfactory results, in achieving our goal. In conclusion, this approach could be well used in reaching our goal based on over 75% of positive feedback (rated very good and good) received. But in order to improve its accuracy and efficiency, more expensive and more accurate distance sensors such as high precision ultrasonic sensors or infrared sensors could be used. The portability, the low cost of engineering, and yet the deliverance of acceptable level of quality of music, could be identified as the unique key point of this research.Item A Simple Machine Learning Approach for Identifying Promotional Short Message Service (SMS) Messages.(Faculty of Computing and Technology, University of Kelaniya, Sri Lanka., 2017) Dias, D.S.; Dias, N.G.J.Mobile phones play an integral part in the modern lives of humans. Short Message Services (SMS) Messages have become a popular mode for simple communication. Its’ simplicity, costeffectiveness and large audience has attracted the attention of advertising industry to send targeted promotional messages to mobile phones. In Sri Lanka, a survey conducted in Colombo, yielded that 3 out of 5 SMS messages received our promotional messages. Even though extensive research has been carried out in detecting junk SMS messages, the amount of research conducted on filtering promotional SMS messages is rare. The purpose of this research is to evaluate the success and accuracy of utilizing a simple machine learning algorithm to identify promotional SMS messages. Here, we have used a feed-forward neural network based on a statistical model, which was trained with a training data set consisting of promotional as well as non-promotional messages. Each test message was broken down in to individual words and filtered through by cleaning to form keywords which will have consist of a weight and probability value. With each message that is used to train, these values will be updated according to whether it is a promotional or a non-promotional message. When a message is tested through this neural network, the words of the message will be matched against the keyword’s weight and probability, which will then calculate a resultant probability. By setting a par-value, we can classify the test as a promotional or a non-promotional message. The proposed model yielded a 100% accuracy when tested within the given test data set. In order to get successful results for broader test data sets, the model has to be trained comprehensively with proper amount of promotional and non-promotional messages. Optionally, the results obtained from the feed forward neural network for incoming messages, can then be fed back in to the feed forward neural network for further training. As future work, we intend to take this solution to an android-based mobile application that extracts promotional messages from the incoming SMS messages as well as from a server, and display them to the user based on his preferences.