Geo-based recommendation system utilising geo tagging and K-means clustering

Amar Shukla,Tanupriya Choudhury,Nehit Benara,Piyush Garg,Aditya Tiwari,Jung‑Sup Um 대한공간정보학회 2023 Spatial Information Research Vol.31 No.3

As technology advances, recommendation systems play an increasingly significant role in everyday life. Users today receive information efficiently and effectively through location-based recommender systems on their mobile devices. Geo-tagged data and the global positioning system are used to gather information about users in location-specific recommender systems. In this busy world, coffee is also a daily requirement. Therefore, we determine whether a particular population of individuals with mobile devices or other utility devices needs recommendations for coffee shops in a particular area. This was achieved by creating a Coffee Shop recommendation system, which uses geotagging to pinpoint the location dependent on latitude and longitude. In this article, we present a machine learning approach to assigning locations to coffee shops based on geo-based location suggestions. To determine the effectiveness of the coffee shop recommendation, a population-based zone-wise analysis was also conducted.

CGI based syslog management system for virtual machines

Kamal Dua,Tanupriya Choudhury,Uppara Rajanikanth,Amitava Choudhury 대한공간정보학회 2020 Spatial Information Research Vol.28 No.4

Undoubtedly, logs are brain of any software system. Development, debugging and upgradation of software applications became much easier due to the logging or auditing concept in computer science field. Virtual Machines also log and timestamp every activity that take place during their runtime. Gathering system details from those lengthy log files is a hectic work for the end-user since single log file contains millennial entries of audited log data. A management system can be very helpful solution and required for the analytical scanning and the visualization of the logged data and thus providing the way to end user for the relevant information about the virtual machines running on the hypervisors. The same system can help developers to gather relevant log entries and presenting data collectively to the end user. This system not only helps in providing and presenting details but also act as an alert system to notify user fatal errors occurred during runtime. Statistical usage information or notification system can further be developed by the means of this syslog system. This paper will present how developers can build an efficient Syslog Management System on the web using the (CGI) Common Gateway Interface in the C programming language and also mentions how CGI environment can be achieved in the Apache Tomcat webserver to build dynamic web tools. Also, it provides basic idea how C can be used effectively in the CGI interface to provide better methods for obtaining and extracting relevant system data.

Spatial optimisation of mango leather production and colour estimation through conventional and novel digital image analysis technique

Sarkar Tanmay,Salauddin Molla,Choudhury Tanupriya,Um Jung-Sup,Pati Siddhartha,Hazra Sudipta Kumar,Chakraborty Runu 대한공간정보학회 2021 Spatial Information Research Vol.29 No.4

Being a seasonal fruit mango cannot be cherished over the year; dehydration may be a solution to preserve the deliquesce of mango as mango leather. The processing parameters like puree load (0.4–0.6 g/cm2 ), total soluble solid (20–30 B), oven temperature (60–80 C), and microwave power level (100–300 W) were optimised for a superior textural attribute (hardness) primitive drying method like sun drying, industrially practiced modern methods like hot air oven drying and microwave drying and cutting-edge drying technique like freeze-drying. Response surface methodology and artificial neural network technique were adapted to model these drying procedures by considering the central composite design. The mathematical operations guiding to describe the model were studied. Being an imperative parameter colour quantification is essential for food industries. Current research employs microwave drying to produce mango leather with colour quantification approach. The L, ‘a’ and ‘b’ values of the product have been measured by Hunter Lab colorimeter and by digital image analysis, to determine the chromatic view harmonious to human vision. The relative analysis of colour measurement through these two techniques has been studied.

An algorithmic approach for performance tuning of a relational database system using dynamic SGA parameters

Sharma Hitesh Kumar,Choudhury Tanupriya,Tomar Ravi,Patni J. C.,Um Jung-Sup 대한공간정보학회 2021 Spatial Information Research Vol.29 No.6

In twenty-first century, spatial database are becoming complex in nature due to the diversity of resources to generate and collect these datasets. Efficient query handing and high performance is the key requirement for success for the expert systems (like GIS) using these spatial databases. However, relational database management systems (RDBMS) used for storing spatial datasets are struggling for auto-tuning, self-diagnosis and self-healing. To overcome these challenges for spatial database and RDBMS, we have identified 250 plus dynamic parameters, which are responsible for managing SGA (system global area) of a running instance of any RDBMS. These parameters can be controlled at runtime and allocation and deallocation of memory to the various components of SGA can be managed by changing the values of these parameters. In this research work, the data is collected related to the system parameters and the resources utilized by the system. The collected data is automatically analyzed to find the relation between parameters and resources. Based upon the relations (direct or inverse) the decision will be taken by the developed utility to enhance overall system performance. In this work a framework, related algorithms, and utility tool to perform above-said steps for improving performance are designed and implemented. This research work will help to get fast and efficient handling of spatial datasets, which will directly affect the performance of expert system to take quick decisions.

Indexing hs code- a hybrid indexer for an optimized search of geotagged data

Bhagwant Singh,Kingshuk Srivastava,D. K. Gupta,Tanupriya Choudhury,Jung‑Sup Um 대한공간정보학회 2023 Spatial Information Research Vol.31 No.1

In case of unstructured data processing current technologies have developed a lot of solutions to process and provide insight into it, which has become paramount today due to the permeation of social media. Social media analytics extract data points by searching the most relevant textual reference and isolating textual data with location information. This increase the complexity of modern social media search engines. The indexer is imperative towards the performance of these engines. Geotagged data generated via modern social media technologies has augmented the need to enhance such search mechanisms designed for spatial data.Conventional spatial indexers designed to handle such data can accurately search spatial objects but with a considerable increase in seek time. This paper presents a hybrid spatial indexer based on “Hs-I” tree for the “Social Media Spatial Analytical (SMSA)” model. The purposed indexer is 17.768% faster when compared with the “Geo-hash” indexer. The model refers “CBDFI” model for base architecture and deploys the advantage of the “hs” code. The paper presents the comparison of the purposed indexer with various other spatial indexers and highlights its key points in terms of execution time.

A roadmap of steganography tools: conventional to modern

Pilania Urmila,Tanwar Rohit,Gupta Prinima,Choudhury Tanupriya 대한공간정보학회 2021 Spatial Information Research Vol.29 No.5

Steganography emerged as an effective technology for securing the data over the network. Its specificity of concealing the existence of the secret data supports its application in securing the information in the modern era. The desire of industry and support from various governments motivated the researchers to develop steganography tools. These spatial and transform domain tools implement different steganography techniques either solo or as a hybrid using a wide range of media as a cover file for hiding various types of data. In this paper, a systematic study of the steganography tools developed in the last three decades has been done. The comparative analysis of these tools based on specified parameters represents their strengths, limitations, applicability, and scope for future work as well. OpenPuff steganography tool spawns a huge acceptance by academics and professionals. This paper also analyze the performance of the OpenPuff tool on some unexplored parameters to validate and justify its performance.

Ascertaining polarity of public opinions on Bangladesh cricket using machine learning techniques

Faruque M. Abdullah,Rahman Saifur,Chakraborty Partha,Choudhury Tanupriya,Um Jung-Sup,Singh Thipendra Pal 대한공간정보학회 2022 Spatial Information Research Vol.30 No.2

In the present world, we are not only the consumers of information but creators as well. The virtual world of social media, which is considered a free open forum for discussion; provides its participants a chance to shape or re-shape the digital information by expressing opinions. These opinions generally contain different types of sentiments. Sentiment analysis is a tool that performs the computational study of identifying and extracting sentiment content of textual data that can be used to classify those public opinions posted on various topics in social media. In this paper, a sentiment polarity detection approach is presented, that detects the polarity of textual Facebook posts in Bangla containing people’s point of views on Bangladesh Cricket using three popular supervised machine learning algorithms named Naive Bayes (NB), support vector machines (SVM), and logistic regression (LR). Comparative result analysis is also provided between classifiers, where LR performed slightly better than SVM and NB by considering n-gram as a feature with an accuracy of 83% in the present world, we are not only the consumers of information but creators as well. The virtual world of social media, which is considered a free open forum for discussion; provides its participants a chance to shape or re-shape the digital information by expressing opinions. These opinions generally contain different types of sentiments. Sentiment analysis is a tool that performs the computational study of identifying and extracting sentiment content of textual data that can be used to classify those public opinions posted on various topics in social media. In this paper, a sentiment polarity detection approach is presented, that detects the polarity of textual Facebook posts in Bangla containing people’s point of views on Bangladesh Cricket using three popular supervised machine learning algorithms named Naive Bayes (NB), support vector machines (SVM), and logistic regression (LR). Comparative result analysis is also provided between classifiers, where LR performed slightly better than SVM and NB by considering n-gram as a feature with an accuracy of 83%.

Combating disaster prone zone by prioritizing attributes with hybrid clustering and ANP approach

Srivastava Rashi,Sabitha Sai,Majumdar Rana,Choudhury Tanupriya,Dewangan Bhupesh Kumar 대한공간정보학회 2021 Spatial Information Research Vol.29 No.4

Due to a lack of assets and high improbability, developing an information system for every disaster situation is a challenging task. Consequently, building an information system for such a situation is essential, and the latest research direction has emphasized on development of such a system which varies due to the irregular nature of environments. This work emphasizes the data mining technique based on available disaster data or the earlier prediction of such occurrences to combat damages. The data mining technique is applied to the clustering of data for smooth processes of the obtained data. K-means clustering and analytic network process (ANP) are implemented as unsupervised learning for initial data and to find groups in the data, clustered based on feature similarity. The proposed approach implies an effective tool for predicting impacts in terms of hazards and this paper also evaluates its effectiveness. This study offers important insights into the disaster recovery practitioner to select the best disaster recovery solution and prioritize them for their enterprise.

Improved email classification through enhanced data preprocessing approach

B. Aruna Kumara,Kodabagi Mallikarjun M.,Choudhury Tanupriya,Um Jung-Sup 대한공간정보학회 2021 Spatial Information Research Vol.29 No.2

Email has become one of the most widely used forms of communication, resulting in an exponential increase in emails received and creating an immense burden on existing approaches to email classification. Applying the classification method on the raw data may worsen the performance of classifier algorithms. Hence, the data have to be prepared for better performance of the machine learning classifiers. This paper proposes an enhanced data preprocessing approach for multi-category email classification. The proposed model removes the signature of the email. Further, special characters and unwanted words are removed using various preprocessing methods such as stop-word removal, enhanced stop-word removal, and stemming. The proposed model is evaluated using various classifiers such as Multi-Nominal Naı¨ve Bayes, Linear Support Vector Classifier, Logistic Regression, and Random Forest. The results showed that the proposed data preprocessing to email classification is superior to the existing approach.

Ascertaining polarity of public opinions on Bangladesh cricket using machine learning techniques

Faruque M. Abdullah,Rahman Saifur,Chakraborty Partha,Choudhury Tanupriya,Um Jung-Sup,Singh Thipendra Pal 대한공간정보학회 2022 Spatial Information Research Vol.30 No.1

In the present world, we are not only the consumers of information but creators as well. The virtual world of social media, which is considered a free open forum for discussion; provides its participants a chance to shape or re-shape the digital information by expressing opinions. These opinions generally contain different types of sentiments. Sentiment analysis is a tool that performs the computational study of identifying and extracting sentiment content of textual data that can be used to classify those public opinions posted on various topics in social media. In this paper, a sentiment polarity detection approach is presented, that detects the polarity of textual Facebook posts in Bangla containing people’s point of views on Bangladesh Cricket using three popular supervised machine learning algorithms named Naive Bayes (NB), support vector machines (SVM), and logistic regression (LR). Comparative result analysis is also provided between classifiers, where LR performed slightly better than SVM and NB by considering n-gram as a feature with an accuracy of 83%.