Browsing by Author "Adikari, S"

Now showing 1 - 5 of 5

item: Conference-Extended-Abstract
Automated taxi dispatching system with accurate travel time prediction
(2014-01-16) Kumarasinghe, S; Adikari, S
Fast and reliable transportation systems demonstrate the wealth of a country. Aim of this project is to develop an integrated system which is capable of addressing both identify patterns of traffic to estimate the travel time for specific route and real-rime update of current incidents on the route. Estimate the travel time for specific destination, identify the shortest time consuming path among optional paths and provide the real time update for route incidents are the main objectives of this automated system. The solution is to develop an automated taxi dispatching system without using operators to communicate with passengers and also providing the travel time to user destination. Travel time prediction with high accuracy is the major task of this system. So the key role of this approach is the predicted speed value of the identified route. This report illustrates the importance of having clear details of longitude, latitude data and other traffic related data for accurate decision making of taxi dispatching. In this prediction process mainly used artificial neural networks to train the database. Pattern recognition leads to identify patterns of traffic related data. This paper presents a novel approach which builds a travel time prediction model based on a historical traffic data and other speed varying factors. Evaluation results shown that the acceptable accuracy of travel speed prediction and how to affect speed varying factors to the error values. This concept will be fruitful for many taxi dispatching systems and fleet management as well as traffic related travel time prediction systems.
item: Conference-Extended-Abstract
Intelligence for web content bookmarking and organization through key-phrase extraction
(2014-01-16) Nilanga, GN; Adikari, S
Main focus of this research is to introduce the novel concept of web content bookmarking, by using a tool called iBookmark. What is meant by 'content bookmarking' in here is, the users of iBookmark tool can easily save the textual content found interesting on web, for later reference in a meaningfully named and organized hierarchical structure. In achieving this, the tool determines a phrase which is suited to label the user selected text by using a technique called 'Key Phrase Extraction'. Those labels are used to create the hierarchy. In brief, 'how to bookmark interesting web content?' is the research question addressed throughout this paper and the novel key phrase extraction mechanism proposed is part of the solution while rest being existing technologies. Throughout this paper, that key phrase extraction technique which is using Artificial Neural Networks is elaborated and it's the major research component in the endeavour to achieve the final goal of content bookmarking. As well, manual selection of key phrases is a human cognition related activity and this tool is automating that task. Hence it's called intelligent bookmarker, iBookmark. Further, this tool is targeting specific user categories such as scholars/researchers since the idea of building such a tool was emerged when thinking how fascinating if such a tool exists to help those specific user categories.
item: Article-Full-text
A new approach to real-time bidding in online advertisements: Auto pricing strategy
(Institute for Operations Research and the Management Sciences, 2019) Adikari, S; Dutta, K
Abstract. Real-time bidding (RTB) for digital advertising is becoming the norm for improving advertisers’ campaigns. Unlike traditional advertising practices, in the process of RTB, the advertisement slots of a mobile application or awebsite are mapped to a particular advertiser through a real-time auction. The auction is triggered and is held for a few milliseconds after an application is launched. As one of the key components of the RTB ecosystem, the demand-side platform gives the advertisers a full pledge window to bid for available impressions. Because of the fast-growing market of mobile applications and websites, the selection of the most pertinent target audience for a particular advertiser is not a simple human-mediated process. The real-time programmatic approach has become popular instead. To address the complexity and dynamic nature of the RTB process, we propose an auto pricing strategy (APS) approach to determine the applications to bid for and their respective bid prices from the advertising agencies’ perspective. We apply the APS to actual RTB data and demonstrate how it outperforms the existing RTB approaches with a higher conversion rate for a lower target spend.
item: Conference-Extended-Abstract
Swarm intelligence for resource allocation of emergency situations in hospitals
(2014-01-16) Rathnayake, H; Adikari, S
This paper focuses on resource scheduling for patients at emergency department in hospitals. Scheduling and coordinating of patients at emergency department is faced with a high amount of complexity due to the inherent dynamics of the processes and the distributed organizational structure of hospitals. As a solution, multi-agent system is presented, in which patients, emergency department human resources and physical resources are represented as autonomous agents. These agents are able to read flexibly to changes and disturbances (e.g. emergencies and complications) through proactiveness and reactivates. The solution is mainly focuses on emergency situations at hospitals. Emergency admission is when admission is unpredictable and at short notice because of clinical need. This paper describes a solution to overcome the difficulties that occurs at the emergency situations. The system has been developed on Jade environment and can run on any compute!".
item: Conference-Full-text
Swarm intelligence for resource allocation of emergency situations in hospitals
(2014-06-24) Rathnayake, HC; Adikari, S
This paper focuses on resource scheduling for patients at emergency department in hospitals. Scheduling and coordinating of patients at emergency department is faced with a high amount of complexity due to the inherent dynamics of the processes and the distributed organizational structure of hospitals. As a solution, multi-agent system is presented, in which patients, emergency department human resources and physical resources are represented as autonomous agents. These agents are able to react flexibly to changes and disturbances (e.g. emergencies and complications) through pro-activeness and reactivates. The solution is mainly focuses on emergency situations at hospitals. Emergency admission is when admission is unpredictable and at short notice because of clinical need. This paper describes a solution to overcome the difficulties that occurs at the emergency situations. The system has been developed on Jade environment and can run on any computer.