Pub Date : 2021-12-06DOI: 10.1109/SLAAI-ICAI54477.2021.9664746
Sulakna Karunaratna, Pasan Maduranga
Recent advances on the Internet of Things (IoT) enable intelligent computing algorithms on tiny hardware devices such as Single Board Computers (SBC). Among popular IoT applications, Sound event recognition and classification have enabled exciting and vital applications. Sound events carry information that is useful for our daily lives. The perception of surrounding events by humans depends strongly on audio signals. Awareness of what happens in the surrounding environment depends heavily on the ability of an individual to perceive sounds and accurately recognize events related to them. This paper presents a study using SBC for machine learning and deep learning-based application and finally evaluates the overall performances against a standard PC and a Raspberry Pi.
{"title":"Artificial Intelligence on Single Board Computers: An Experiment on Sound Event Classification","authors":"Sulakna Karunaratna, Pasan Maduranga","doi":"10.1109/SLAAI-ICAI54477.2021.9664746","DOIUrl":"https://doi.org/10.1109/SLAAI-ICAI54477.2021.9664746","url":null,"abstract":"Recent advances on the Internet of Things (IoT) enable intelligent computing algorithms on tiny hardware devices such as Single Board Computers (SBC). Among popular IoT applications, Sound event recognition and classification have enabled exciting and vital applications. Sound events carry information that is useful for our daily lives. The perception of surrounding events by humans depends strongly on audio signals. Awareness of what happens in the surrounding environment depends heavily on the ability of an individual to perceive sounds and accurately recognize events related to them. This paper presents a study using SBC for machine learning and deep learning-based application and finally evaluates the overall performances against a standard PC and a Raspberry Pi.","PeriodicalId":252006,"journal":{"name":"2021 5th SLAAI International Conference on Artificial Intelligence (SLAAI-ICAI)","volume":"22 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133725592","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-12-06DOI: 10.1109/SLAAI-ICAI54477.2021.9664718
Hansika Gunasekara, Thushari P. Silva
The rapid growth of the web and its applications has created immense importance for recommender systems. Recommender systems were designed to generate suggestions for items or services based on user interests with the applications to different domains. However, the integration of multiple data sources while resolving semantic ambiguity of entities involved in the integration has been overlooked in many recommender systems developed for travel recommendation. This research proposes an ontology-based travel recommender system to overcome such deficiencies in the current travel recommender systems. The developed ontology facilitates the integration of multi-model data for personalized travel recommendations. The similarity analysis of entities to be interconnected is performed by using a semantic data classification technique that integrates a hybrid filtering approach to classify similar entities, including tours and visitors. The proposed ontology-based approach for travel recommendation outperforms other methods and with higher accuracies.
{"title":"Personalized Travel Recommendation System Using an Ontology","authors":"Hansika Gunasekara, Thushari P. Silva","doi":"10.1109/SLAAI-ICAI54477.2021.9664718","DOIUrl":"https://doi.org/10.1109/SLAAI-ICAI54477.2021.9664718","url":null,"abstract":"The rapid growth of the web and its applications has created immense importance for recommender systems. Recommender systems were designed to generate suggestions for items or services based on user interests with the applications to different domains. However, the integration of multiple data sources while resolving semantic ambiguity of entities involved in the integration has been overlooked in many recommender systems developed for travel recommendation. This research proposes an ontology-based travel recommender system to overcome such deficiencies in the current travel recommender systems. The developed ontology facilitates the integration of multi-model data for personalized travel recommendations. The similarity analysis of entities to be interconnected is performed by using a semantic data classification technique that integrates a hybrid filtering approach to classify similar entities, including tours and visitors. The proposed ontology-based approach for travel recommendation outperforms other methods and with higher accuracies.","PeriodicalId":252006,"journal":{"name":"2021 5th SLAAI International Conference on Artificial Intelligence (SLAAI-ICAI)","volume":"15 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134633889","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-12-06DOI: 10.1109/SLAAI-ICAI54477.2021.9664678
Heshani Bopage, Chinthanie Weerakoon
Relaxation is the emotional state of a living being, of low tension, in which there is an absence of arousal that could come from sources such as anger, anxiety, or fear. Technology can be used to mind relaxation. Chatbot is an automated computer software program capable of having intelligent live conversations with people. It is a technology that provides a new way to interact with computer systems. Nowadays, chatbots are very popular in a large scale of applications, especially in systems that provide intelligence support to the user. It is one of the technologies that has been used successfully in many fields such as education, marketing and health field. This paper aims to develop such a chatbot for the mind relaxation of university students using Natural Language processing techniques and Dense Neural Network. The chatbot tool was trained with a series of counseling conversations in text form. Training phases included intents, tags, patterns and responses. Primary function of chatbot have to play is to understand the intents of students and to respond to them appropriately. Input and Output are in the form of the text.
{"title":"Mind Relaxation Chatbot for University Students by Using Dense Neural Network","authors":"Heshani Bopage, Chinthanie Weerakoon","doi":"10.1109/SLAAI-ICAI54477.2021.9664678","DOIUrl":"https://doi.org/10.1109/SLAAI-ICAI54477.2021.9664678","url":null,"abstract":"Relaxation is the emotional state of a living being, of low tension, in which there is an absence of arousal that could come from sources such as anger, anxiety, or fear. Technology can be used to mind relaxation. Chatbot is an automated computer software program capable of having intelligent live conversations with people. It is a technology that provides a new way to interact with computer systems. Nowadays, chatbots are very popular in a large scale of applications, especially in systems that provide intelligence support to the user. It is one of the technologies that has been used successfully in many fields such as education, marketing and health field. This paper aims to develop such a chatbot for the mind relaxation of university students using Natural Language processing techniques and Dense Neural Network. The chatbot tool was trained with a series of counseling conversations in text form. Training phases included intents, tags, patterns and responses. Primary function of chatbot have to play is to understand the intents of students and to respond to them appropriately. Input and Output are in the form of the text.","PeriodicalId":252006,"journal":{"name":"2021 5th SLAAI International Conference on Artificial Intelligence (SLAAI-ICAI)","volume":"244 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131898028","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-12-06DOI: 10.1109/SLAAI-ICAI54477.2021.9664660
Madhuka D. Bandara, Sandeepa Viduranga, Nipun Rodrigo, Menaka Ranasinghe
Systems integrated with the power of artificial Intelligence and embodied in real environments are having the power of operating autonomously and is capable of performing real-time activities with emotional intelligence. Such systems are simply known as artificial cognitive systems embodied in real environments. These systems are capable of taking sensory inputs process them and act in the environment with cognitive capabilities. The cognitive capabilities of these systems are totally dependent on the architecture adopted in these systems. These cognitive architectures provide an abstract perception model for these cognitive systems. These cognitive architectures are inspired by how physical systems operates autonomously by the combination of brain, mind and body combination thus obligating the necessity of investigating system embodiment considering physical computational properties as well as non-representational properties. Hence this paper investigates the existing architectures for embodiment. It is noted that achieving complete embodiment of systems on par with humans poses a research challenge and is being extensively investigated. Through the literature review it is identified that the use of emergent architectures combined with hybrid approaches as a feasible way of achieving embodiment in artificial cognitive systems.
{"title":"Architectures used in Artificial Cognitive Systems for Embodiment","authors":"Madhuka D. Bandara, Sandeepa Viduranga, Nipun Rodrigo, Menaka Ranasinghe","doi":"10.1109/SLAAI-ICAI54477.2021.9664660","DOIUrl":"https://doi.org/10.1109/SLAAI-ICAI54477.2021.9664660","url":null,"abstract":"Systems integrated with the power of artificial Intelligence and embodied in real environments are having the power of operating autonomously and is capable of performing real-time activities with emotional intelligence. Such systems are simply known as artificial cognitive systems embodied in real environments. These systems are capable of taking sensory inputs process them and act in the environment with cognitive capabilities. The cognitive capabilities of these systems are totally dependent on the architecture adopted in these systems. These cognitive architectures provide an abstract perception model for these cognitive systems. These cognitive architectures are inspired by how physical systems operates autonomously by the combination of brain, mind and body combination thus obligating the necessity of investigating system embodiment considering physical computational properties as well as non-representational properties. Hence this paper investigates the existing architectures for embodiment. It is noted that achieving complete embodiment of systems on par with humans poses a research challenge and is being extensively investigated. Through the literature review it is identified that the use of emergent architectures combined with hybrid approaches as a feasible way of achieving embodiment in artificial cognitive systems.","PeriodicalId":252006,"journal":{"name":"2021 5th SLAAI International Conference on Artificial Intelligence (SLAAI-ICAI)","volume":"39 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124050867","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-12-06DOI: 10.1109/SLAAI-ICAI54477.2021.9664736
R. Ranawana, A. Karunananda
Software development teams are often hampered when aligning machine learning production with standard software development processes. Iterative experimentation is needed to address the inherent complexities of data collection and preparation, model entanglement, and the technical debt of machine learning. The complexity of this process is compounded due to dependencies on the production environment and real- time data. We propose a unified framework which facilitates the planning, development, and deployment of a machine learning application through parallel processes for software and machine learning engineering. This allows for the risk of both the project and machine learning development to be significantly reduced through continuous integration, evaluation, and production. The framework, named MLASDLC, unifies concepts from standard software development life cycle methodologies (SDLC), development operations (DevOps) and machine learning operations (MLOps) to present a framework for the development of machine learning applications.
{"title":"An Agile Software Development Life Cycle Model for Machine Learning Application Development","authors":"R. Ranawana, A. Karunananda","doi":"10.1109/SLAAI-ICAI54477.2021.9664736","DOIUrl":"https://doi.org/10.1109/SLAAI-ICAI54477.2021.9664736","url":null,"abstract":"Software development teams are often hampered when aligning machine learning production with standard software development processes. Iterative experimentation is needed to address the inherent complexities of data collection and preparation, model entanglement, and the technical debt of machine learning. The complexity of this process is compounded due to dependencies on the production environment and real- time data. We propose a unified framework which facilitates the planning, development, and deployment of a machine learning application through parallel processes for software and machine learning engineering. This allows for the risk of both the project and machine learning development to be significantly reduced through continuous integration, evaluation, and production. The framework, named MLASDLC, unifies concepts from standard software development life cycle methodologies (SDLC), development operations (DevOps) and machine learning operations (MLOps) to present a framework for the development of machine learning applications.","PeriodicalId":252006,"journal":{"name":"2021 5th SLAAI International Conference on Artificial Intelligence (SLAAI-ICAI)","volume":"7 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127518436","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-12-06DOI: 10.1109/SLAAI-ICAI54477.2021.9664748
D. Vidanagama, Thushari P. Silva, A. Karunananda
Nowadays, due to the prevailing situation of the world, people are heavily focusing on online transactions. There has been a rapid increase in online transactions and several types of data generated through such transactions during the last few years. As there is no other involvement in purchasing decisions, customers make purchasing judgments through the reviews. Therefore, not only for making purchasing decisions but also customer reviews provide valuable information regarding the products for decision-makers. By considering this as an advantage, fraudulent reviewers tend to write reviews to promote or downgrade products. Deceptive reviews can be identified via reviewer behavioural features, content-related features, or review features. But all the extracted features may not be critical for identifying deceptive. This research introduces a novel filter-wrapper hybrid approach to select optimal features to identify deceptive online customer reviews. A combination of univariate and multivariate filter methods as well as a wrapper method with the bidirectional search were used to select the features. The model was evaluated using the K-Nearest Neighbor (KNN) classifier. The proposed hybrid approach shows the highest model accuracy against the sole traditional approaches. The selected optimal features used for model building are effective as they reveal the most statistically significant features when predicting the deceptive reviews..
{"title":"Hybrid Filter-Wrapper Approach for Feature Selection in Deceptive Consumer Review Classification","authors":"D. Vidanagama, Thushari P. Silva, A. Karunananda","doi":"10.1109/SLAAI-ICAI54477.2021.9664748","DOIUrl":"https://doi.org/10.1109/SLAAI-ICAI54477.2021.9664748","url":null,"abstract":"Nowadays, due to the prevailing situation of the world, people are heavily focusing on online transactions. There has been a rapid increase in online transactions and several types of data generated through such transactions during the last few years. As there is no other involvement in purchasing decisions, customers make purchasing judgments through the reviews. Therefore, not only for making purchasing decisions but also customer reviews provide valuable information regarding the products for decision-makers. By considering this as an advantage, fraudulent reviewers tend to write reviews to promote or downgrade products. Deceptive reviews can be identified via reviewer behavioural features, content-related features, or review features. But all the extracted features may not be critical for identifying deceptive. This research introduces a novel filter-wrapper hybrid approach to select optimal features to identify deceptive online customer reviews. A combination of univariate and multivariate filter methods as well as a wrapper method with the bidirectional search were used to select the features. The model was evaluated using the K-Nearest Neighbor (KNN) classifier. The proposed hybrid approach shows the highest model accuracy against the sole traditional approaches. The selected optimal features used for model building are effective as they reveal the most statistically significant features when predicting the deceptive reviews..","PeriodicalId":252006,"journal":{"name":"2021 5th SLAAI International Conference on Artificial Intelligence (SLAAI-ICAI)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121091804","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-12-06DOI: 10.1109/SLAAI-ICAI54477.2021.9664742
Chamali Basnayake, C. Peiris, H. Wickramarathna, Poornima Jayathunga
In the modern world, professionals of diverse industrial sectors have severely become victims of Obese and Overweight conditions. Obese and Overweight conditions can be minimized by having proper dietary plans, physical activities and minimizing alcohol-based relaxation. In this research context, we try to address the issue of having poor physical exercise. We guide professionals with suitable exercises to reduce their weight in order to have the required Body Mass Index(BMI). The user body measurements that are recommended by the domain experts to concern such as sex, height, weight, exercise preferences, age, diet details and medical history used to calculate the degree of obesity of each individual. Then the degree of obesity is mapped with the knowledge base along with the predefined rules in order to match respective exercises suitable for the particular individuals that are compatible with the user’s medical history. Two ontologies for foods and exercises were developed using Protégé 4.3.0 and were retrieved by running Simple Protocol and Resource Description Framework Query Language (SPARQL) queries. Python 3 is used as the backend language for ontology and interface integration. Frontend developed using Tkinter GUI in Python 3 and is presented for the users to ease the interaction with the system. Two ontological files of Foods and Exercises are loaded and tested for consistency using the HermiT reasoner with the aid of Owlready2. Accuracy and Correctness are checked by addressing the competency questions and by domain experts’ inspections.
{"title":"Recommender System based on Food and Exercise Ontologies to Find the Suitable Fitness Exercise Plan with the Aid of Python","authors":"Chamali Basnayake, C. Peiris, H. Wickramarathna, Poornima Jayathunga","doi":"10.1109/SLAAI-ICAI54477.2021.9664742","DOIUrl":"https://doi.org/10.1109/SLAAI-ICAI54477.2021.9664742","url":null,"abstract":"In the modern world, professionals of diverse industrial sectors have severely become victims of Obese and Overweight conditions. Obese and Overweight conditions can be minimized by having proper dietary plans, physical activities and minimizing alcohol-based relaxation. In this research context, we try to address the issue of having poor physical exercise. We guide professionals with suitable exercises to reduce their weight in order to have the required Body Mass Index(BMI). The user body measurements that are recommended by the domain experts to concern such as sex, height, weight, exercise preferences, age, diet details and medical history used to calculate the degree of obesity of each individual. Then the degree of obesity is mapped with the knowledge base along with the predefined rules in order to match respective exercises suitable for the particular individuals that are compatible with the user’s medical history. Two ontologies for foods and exercises were developed using Protégé 4.3.0 and were retrieved by running Simple Protocol and Resource Description Framework Query Language (SPARQL) queries. Python 3 is used as the backend language for ontology and interface integration. Frontend developed using Tkinter GUI in Python 3 and is presented for the users to ease the interaction with the system. Two ontological files of Foods and Exercises are loaded and tested for consistency using the HermiT reasoner with the aid of Owlready2. Accuracy and Correctness are checked by addressing the competency questions and by domain experts’ inspections.","PeriodicalId":252006,"journal":{"name":"2021 5th SLAAI International Conference on Artificial Intelligence (SLAAI-ICAI)","volume":"61 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116713858","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-12-06DOI: 10.1109/SLAAI-ICAI54477.2021.9664707
Chuangmin Xie, Degang Chen, Hao Shi, Mingyu Fan
Relation Extraction (RE) is one of the most important tasks in Natural Language Processing (NLP). In recent years, with the development of deep learning, a variety of deep neural networks, such as Convolution Neural Network (CNN), Recurrent Neural Network (RNN) and Long Short Term Memory Network (LSTM), have been used in relation extraction and made significant progress. Moreover, LSTM has become the mainstream model in the field of NLP due to its better long term dependencies capture capability than CNN. However, the ability of LSTM to capture long term dependencies is still limited. In order to solve this problem, we propose a phrase convolution structure. The structure can extract the phrase-level features of the sentence, and the sentence-level features can be further extracted after the features are input into LSTM. We believe that this actually enhances the ability of LSTM to capture long term dependencies. Our experiments on SemEva1-2010 Task 8 dataset show that the performance of our model is better than most existing models.
{"title":"Attention-Based Bidirectional Long Short Term Memory Networks Combine with Phrase Convolution Layer for Relation Extraction","authors":"Chuangmin Xie, Degang Chen, Hao Shi, Mingyu Fan","doi":"10.1109/SLAAI-ICAI54477.2021.9664707","DOIUrl":"https://doi.org/10.1109/SLAAI-ICAI54477.2021.9664707","url":null,"abstract":"Relation Extraction (RE) is one of the most important tasks in Natural Language Processing (NLP). In recent years, with the development of deep learning, a variety of deep neural networks, such as Convolution Neural Network (CNN), Recurrent Neural Network (RNN) and Long Short Term Memory Network (LSTM), have been used in relation extraction and made significant progress. Moreover, LSTM has become the mainstream model in the field of NLP due to its better long term dependencies capture capability than CNN. However, the ability of LSTM to capture long term dependencies is still limited. In order to solve this problem, we propose a phrase convolution structure. The structure can extract the phrase-level features of the sentence, and the sentence-level features can be further extracted after the features are input into LSTM. We believe that this actually enhances the ability of LSTM to capture long term dependencies. Our experiments on SemEva1-2010 Task 8 dataset show that the performance of our model is better than most existing models.","PeriodicalId":252006,"journal":{"name":"2021 5th SLAAI International Conference on Artificial Intelligence (SLAAI-ICAI)","volume":"129 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123407740","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-12-06DOI: 10.1109/SLAAI-ICAI54477.2021.9664714
A. Karunananda, Thushari P. Silva, Dewmal Handapangoda, Sadika Sumanapala
Research is popularly known as a process. Many undergraduate, master and even doctoral students find it challenging to go through the research process comfortably. In this sense, novice researchers find pitfalls in the systematic literature review, research problem definition, methodology formulation, and evaluation. At present, some software solutions are available for supporting only specific steps such as literature review, data collection and data analysis. In filling this gap, we have developed an Intelligent Personal Research Assistant, InPRA, which can support the complete research process and scientific writing. InPRA has been powered by language processing, semantic analysis based on topic modelling, and machine learning in artificial intelligence. Based on the profile of a research student, InPRA recommends literature, guide through a systematic literature review leading to the definition of a research problem, formulating a methodology and executing it to generate a conclusion based on an evaluation. InPRA has a distinct feature to guide through writing papers and a thesis incrementally while conducting the research. Based on the evaluation results, the relatedness of the articles generated by InPRA is higher than traditional keyword-based searching. A supervisor can also use this software to monitor and keep track of the progress of their research student.
{"title":"Intelligent Personal Research Assistant","authors":"A. Karunananda, Thushari P. Silva, Dewmal Handapangoda, Sadika Sumanapala","doi":"10.1109/SLAAI-ICAI54477.2021.9664714","DOIUrl":"https://doi.org/10.1109/SLAAI-ICAI54477.2021.9664714","url":null,"abstract":"Research is popularly known as a process. Many undergraduate, master and even doctoral students find it challenging to go through the research process comfortably. In this sense, novice researchers find pitfalls in the systematic literature review, research problem definition, methodology formulation, and evaluation. At present, some software solutions are available for supporting only specific steps such as literature review, data collection and data analysis. In filling this gap, we have developed an Intelligent Personal Research Assistant, InPRA, which can support the complete research process and scientific writing. InPRA has been powered by language processing, semantic analysis based on topic modelling, and machine learning in artificial intelligence. Based on the profile of a research student, InPRA recommends literature, guide through a systematic literature review leading to the definition of a research problem, formulating a methodology and executing it to generate a conclusion based on an evaluation. InPRA has a distinct feature to guide through writing papers and a thesis incrementally while conducting the research. Based on the evaluation results, the relatedness of the articles generated by InPRA is higher than traditional keyword-based searching. A supervisor can also use this software to monitor and keep track of the progress of their research student.","PeriodicalId":252006,"journal":{"name":"2021 5th SLAAI International Conference on Artificial Intelligence (SLAAI-ICAI)","volume":"55 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123749577","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2021-12-06DOI: 10.1109/SLAAI-ICAI54477.2021.9664708
Prabodha Chandrapala, N. Yapage, Meril Mendis
The National Meteorological Center, Department of Meteorology, Sri Lanka is not currently using technologically advanced methods in forecasting daily minimum and maximum temperature of selected locations in the country. In the city weather forecast, they mainly focus on ten cities namely, Anuradhapura, Badulla, Batticaloa, Colombo, Galle, Hambantota, Jaffna, Kandy, Ratnapura, and Trincomalee, covering the entire island. Motivated by the requirement for a sophisticated forecasting technique, we introduce an Artificial Neural Network (ANN) approach for this problem using previous weather data as inputs from more than ten locations in Sri Lanka over ten years (2010-2019). The data used in this work were obtained from the Department of Meteorology, Sri Lanka. A three-layer (input, hidden and output) ANN having appropriate number of nodes in each layer and with the Ward architecture was constructed which uses three activation functions (Gaussian, Gaussian complement, and hyperbolic tangent) in the hidden layer. The model was validated using the k-fold cross-validation procedure. The results, that is, daily minimum and maximum temperature, were obtained using the R software package (4.0.3 version). It was observed that the predicted values were very homogeneous compared to the real values with a small error and this error was reduced using the gradient descent method. We further investigated how various choices of the number of hidden neurons and the epochs affect these results. It was found that the best number of neurons in the hidden layer was twenty one and if the number of epochs was increased the error was approaching zero. A close agreement between the real and predicted temperature values were observed in this work.
{"title":"Daily Minimum and Maximum Temperature Forecasting in Sri Lanka: An Artificial Neural Network Approach","authors":"Prabodha Chandrapala, N. Yapage, Meril Mendis","doi":"10.1109/SLAAI-ICAI54477.2021.9664708","DOIUrl":"https://doi.org/10.1109/SLAAI-ICAI54477.2021.9664708","url":null,"abstract":"The National Meteorological Center, Department of Meteorology, Sri Lanka is not currently using technologically advanced methods in forecasting daily minimum and maximum temperature of selected locations in the country. In the city weather forecast, they mainly focus on ten cities namely, Anuradhapura, Badulla, Batticaloa, Colombo, Galle, Hambantota, Jaffna, Kandy, Ratnapura, and Trincomalee, covering the entire island. Motivated by the requirement for a sophisticated forecasting technique, we introduce an Artificial Neural Network (ANN) approach for this problem using previous weather data as inputs from more than ten locations in Sri Lanka over ten years (2010-2019). The data used in this work were obtained from the Department of Meteorology, Sri Lanka. A three-layer (input, hidden and output) ANN having appropriate number of nodes in each layer and with the Ward architecture was constructed which uses three activation functions (Gaussian, Gaussian complement, and hyperbolic tangent) in the hidden layer. The model was validated using the k-fold cross-validation procedure. The results, that is, daily minimum and maximum temperature, were obtained using the R software package (4.0.3 version). It was observed that the predicted values were very homogeneous compared to the real values with a small error and this error was reduced using the gradient descent method. We further investigated how various choices of the number of hidden neurons and the epochs affect these results. It was found that the best number of neurons in the hidden layer was twenty one and if the number of epochs was increased the error was approaching zero. A close agreement between the real and predicted temperature values were observed in this work.","PeriodicalId":252006,"journal":{"name":"2021 5th SLAAI International Conference on Artificial Intelligence (SLAAI-ICAI)","volume":"36 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2021-12-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132319666","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：