Dr. Partha Ghosh
@aot.edu.in
Assistant Professor, Department of Computer Science and Engineering
Academy of Technology
Dr. Partha Ghosh is working as an Assistant Professor in the Department of Computer Science and Engineering at Academy of Technology under Maulana Abul Kalam Azad University of Technology, West Bengal, India. He completed his Ph.D. (Tech.) in Information Technology in the year of 2023 from the University of Calcutta, Kolkata, India. Prior to this, Dr. Ghosh completed his M.Tech. in 2013 from the same university. He has published 26 research articles in various reputed peer-reviewed journals, international conferences, and book chapters. Dr. Ghosh has served as a reviewer for numerous international conferences and has been a member of the technical program committees for several international conferences. Additionally, he is a lifetime member of IETE. His research areas include Optimization Techniques, Data Warehousing, Big Data Analysis, Machine Learning, and Business Intelligence.
EDUCATION
M.Tech. and Ph.D (Tech.) from University of Calcutta.
RESEARCH, TEACHING, or OTHER INTERESTS
Computer Science, Artificial Intelligence, Computer Science Applications
Scopus Publications
Scholar Citations
Scholar h-index
Scholar i10-index
Scopus Publications
Partha Ghosh, Oendrila Samanta, Takaaki Goto, and Soumya Sen
Institute of Electrical and Electronics Engineers (IEEE)
Enhancement of the profitability of any business organization is driven by proper forecasting. However, this is challenging as many factors affect the forecasting and the degree of relevant factors changes over time. Henceforth, it is essential for any business organization to develop a reliable and consistent sales forecasting model that can drive their growth. In today’s business environment, customer ratings play a pivotal role in evaluating business performance, particularly in online retailing. These ratings provide valuable insights into the strengths and weaknesses of a product or service. The rating values are generally a set of integer values within a given range. This policy restricts users from expressing their views as they may wish to give a value that is not an integer. Hence, the system fails to capture the actual view of the customer about a certain product or service. As the intermediate values (decimal values) are not permitted, customers are generally compelled to round up their ratings, resulting overrating products. This problem can be addressed if textual reviews from the customers are recorded and these are analyzed for judging customers’ satisfaction level. In this research work, we compute customer satisfaction by analyzing the review text of each customer for a particular product by using VADER sentiment analysis tool and use this result for tuning the actual user given ratings. A novel model is proposed to consider the tuned average customer rating amalgamating with standard forecasting methods like ARIMA, SARIMA, and LSTM. The experimental results on the Amazon dataset reveal 10% to 96% improvement in forecasted values for different types of products.
Partha Ghosh, Takaaki Goto, Leena Jana Ghosh, Giridhar Maji, and Soumya Sen
IGI Global
In the countries or areas where the supply-demand ratio of blood is not maintained, the medication process is being deteriorated, and this may be as fatal as death of the patients. It is being observed in different areas in different seasons or may be at the time of festival scarcity of blood may happen. On the other hand, if the blood donation camp is organized frequently, there may be a surplus of blood as it has expiry dates. Along with these issues, due to the transportation or mismanagement, blood units are wasted. These problems are addressed in this research work, and methodologies are proposed to determine the most suitable blood bank with respect to the blood donation camp. Further, a demand forecasting algorithm is used both for predicting the blood unit demand of every blood bank and for transferring excess blood units to the blood bank where it is needed the most, and also, for the efficient transportation of the blood units, taxicab geometry-based paths are employed.
Partha Ghosh, Takaaki Goto, Leena Jana Ghosh, and Soumya Sen
IEEE
Blood is the indispensable circulating fluid for sustaining human life. On demand supply of quality blood is a big challenge for every government in all developing countries. Specially, in festive seasons and winter, supplying quality blood on time is a big medical challenge. On the other hand, the consequences of mismanaged blood donation camp may lead to excess supply of human blood units. Also, in some cases, it is being noticed that human blood units are getting corrupted in transit from the blood donation camp to the blood bank. Hence, several units of human blood are getting spoiled over the time due to mismanagement and/or maintenance. In this research, we have applied a lexicographic optimization based model for finding best available blood bank from the point of blood donation camp. Alternative taxicab geometry based paths are used for finding best possible shortest path from the blood donation camp to the blood bank.
Zhang Xiao, Takaaki Goto, Partha Ghosh, Tadaaki Kirishima, and Kensei Tsuchida
IEEE
Detection of novel game cheating tools is critical for ensuring fair online play. Such cheating tools are visual-based and effectively avoid detection because they do not change the data of game software. With the development and popularity of artificial intelligence technology, it has become easier for individuals to develop cheating tools, such as a new cheating tool for first-person shooter games that searches for characters on the game screen and automatically targets them. Therefore, in this study, a new cheat detection method is proposed using machine learning. The proposed method can be used to detect new cheating tools based on object detection.
Anjan Dutta, Janapriyo Maitra, Partha Ghosh, Punyasha Chatterjee, Takaaki Goto, and Soumya Sen
ACM
In the last decade, huge growth is observed in the online marketplace. Consumers are increasingly engaged in online shopping due to its operational flexibility, huge product search space and diversified products. In this virtual marketplace, consumers make the purchase decision not only based on product specifications given by the seller but also considers the product reviews given by the peer customers. On the other hand, inconsistent reviews make the consumers confused and thus have a negative impact on the overall sales of a business entity. Therefore, identifying the reviewers giving precise and consistent opinions about the products is important for business organizations. In this work, a novel methodology is suggested to classify reviewers into different categories based on the preciseness of the reviews. Business entities can utilize the precise reviewers for getting genuine product feedback and encourage this group to promote their products resulting in increased business growth.
Partha Ghosh, Takaaki Goto, J. K. Mandal, and Soumya Sen
Springer Science and Business Media LLC
Partha Ghosh, Takaaki Goto, and Soumya Sen
Springer Singapore
Partha Ghosh, Deep Sadhu, and Soumya Sen
Zarqa University
Partha Ghosh, Takaaki Goto, Jyotsna Kumar Mandal, and Soumya Sen
ACM
Data warehouse is used across the organizations for analytical processing. It is organized in the form of lattice of cuboids for carrying out the different types of analysis. The structure of lattice of cuboids grows exponentially with the number of dimensions. Every cuboid contains huge amount of data. The data size is multiplied in the data warehouse as a new entry takes place in any dimension. In modern day business analysis, processing the data in real-time is one of the intrinsic requirements. However, with the gigantic structure of lattice of cuboids along with the highly loaded data in every cuboid it is difficult to achieve the real-time processing. View materialization is practiced over the decades for faster query processing in the database applications. This research proposed a fuzzy based materialized data-cube driven warehouse architecture for fast decision making. Experimental output shows the proposed methodology achieves faster data analysis and better hit-miss ratio in the materialized data-cubes.
Partha Ghosh, Soumya Sen, and Agostino Cortesi
Springer Science and Business Media LLC
Partha Ghosh, Leena Jana Ghosh, Subhajit Guha, Narayan C. Debnath, and Soumya Sen
Springer Singapore
Sinthia Roy, Arijit Banerjee, Partha Ghosh, Amlan Chatterjee, and Soumya Sen
Springer Singapore
Lalmohan Dutta, Giridhar Maji, Partha Ghosh, and Soumya Sen
Springer Singapore
Partha Ghosh, Takaaki Goto, and Soumya Sen
IGI Global
This article describes how multi-criteria decision making problems are difficult to handle in normal SQL query processing. Skyline computation is generally used to solve these types of requirements by using dominance analysis and finding shortest distance with respect to a prime interesting point. However, in real life scenarios shortest distance may not be applicable in most of the cases due to different obstacles or barriers exist between the point of interests or places. In order to consider the presence of obstacles for geographically dispersed data, this research work uses Taxicab geometry for distance calculation, which is a simple Non-Euclidian geometry with minimum time complexity. Another limitation of previous skyline based works are that they only focus upon a single interesting point and can't be apply for multiple interesting points. This research article focuses upon multiple visiting points for the travelers in an optimized way. In addition to this, the article also selects areas for setting up of new business properties considering the constraints.
Partha Ghosh, Subhranil Som, and Soumya Sen
IEEE
Virtual Data warehouses (VDW) are generally used to store and analyze rapidly generated data and thereafter to make quick decision. VDW may stores data in a schema-less form and without any pre-processing to make the decision making on the fly. But for long term and effective business report generation, VDWs must be linked with physical data warehouse. In real life it is often found that a particular attribute is present in VDWs that is missing in actual data warehouse. In this type of situation the system needs to decide whether the attribute will be added in to physical data warehouse or not. Also in many time it is found that a particular attribute has a poor priority for data analysis, but it found very importance in VDWs. To handle all these, alone VDWs or client ends are not fit enough as they don’t have all the information about the system. Hence a specially designated node is required for gathering all the information about the system and changing the structure of the data warehouse.
Partha Ghosh, Takaaki Goto, and Soumya Sen
IEEE
Skyline computation is relevant in multi-criteria decision making where the criteria are inversely proportional to each other. Skyline is generally computed using dominance analysis and applicable in a situation where shortest distance is computed with respect to a point of importance. In real life scenarios different cost parameters are obviously high for the points which are designated as "important" where as users search for the points which are generally of low cost. These types of inverse conditions are managed in skyline computation. Existing research works majorly apply shortest distance calculation for searching the points of importance and it is assumed that the points are connected without any obstructions. However in practical cases this assumption is often wrong as different obstacles or barriers exist between the points or places. In this research work we use Taxicab distance calculation to consider the presence of obstacles and apply it to compute skyline of geographically dispersed data.
Partha Ghosh and Soumya Sen
IEEE
Single-criteria decision making queries can be answered using simple SQL queries, however a multi-criteria decision making problems are often not answered by normal SQL queries. In order to solve these types of queries we may need to use co-operative query languages etc. However using additional query based system incurs extra cost. Moreover, if the criteria in a query are complementary to each other simple SQL queries are not capable of addressing this issue. A query in which multi-criteria decision making is required, often more than a single attribute of the relation is analyzed to fetch the desired result. In this context dominance analysis is performed to obtain a set of points (tuples) those are at least equally good in all the dimensions in compare to other points in the dataset. Skyline points are computed to find points which are not dominated (dominance analysis) by any other point in the system. A point is called “skyline point” if and only if it is not dominated by any other points in the system. Computation of skyline requires comparison of each point to all the other points in the system which in turn increases complexity. The complexity may increase at exponential rate when the numbers of dimensions increase. This research work focuses on the reduction of computational complexity. It is incorporated here by selecting the most important dimension of the database and transfers the other entire dimension in that form. And finally ranks the points accordingly.
Partha Ghosh and Soumya Sen
IEEE
In a multi-criteria decision making system it is required to analyze more than one attribute of the table. In this context no one would be interested on a point (tuples) which is dominated by any other point in the system, they are only interested on skyline points. A point is called a skyline point if and only if it is not dominated by any other point in the system. Therefore for the multi-criteria decision making system, it needs to compute the points those are not dominated by any other point in the system. And then find the best suited value using some mathematical operations, and finally display the list of that in the sorted order, starting from the best value. This paper propose a effective way of computing skyline and ranking points by finding the Best Skyline Point (BSP) among the dataset.
Soumya Sen, Partha Ghosh, and Agostino Cortesi
Springer India
Partha Ghosh and Soumya Sen
IEEE
View materialization is being practiced over several years in large data centric applications like database, data warehouse, data mining etc. for faster query processing. Initially the materialized views are formed based on some methodologies, however the performance (hit-miss ratio) of the materialized views may degrade after certain time if the incoming query pattern changes. This situation could be handled efficiently by employing a view maintenance scheme which works dynamically during query execution at run time. As these materialized views involves huge amount of data, consideration of time and space complexity during the maintenance process plays an important role. In this paper authors adopt an incremental view maintenance policy based on attribute affinity to update the materialized views at run time without using extra space and minimizing the data transfer between the secondary memory and primary memory (where the active materialized views reside). This in turn reduces time complexity and supports incremental maintenance eliminating the requirement of full replacement of existing materialized views.
Partha Ghosh and Soumya Sen
IEEE
Materialized view is used in large data centric applications to expedite query processing. The efficiency of materialized view depends on degree of result found against the queries over the existing materialized views. Materialized views are constructed following different methodologies. Thus the efficacy of the materialized views depends on the methodology based on which these are formed. Construction of materialized views are often time consuming and moreover after a certain time the performance of the materialized views degrade when the nature of queries change. In this situation either new materialized views could be constructed from scratch or the existing views could be upgraded. Fresh construction of materialized views has higher time complexity hence the modification of the existing views is a better solution. Modification process of materialized view is classified under materialized view maintenance scheme. Materialized view maintenance is a continuous process and the system could be tuned to ensure a constant rate of performance. If a materialized view construction process is not supported by materialized view maintenance scheme that system would suffer from performance degradation. In this paper a new materialized view maintenance scheme is proposed using Markov's analysis to ensure consistent performance. Markov's analysis is chosen here to predict steady state probability over initial probability.
Partha Ghosh and Soumya Sen
IEEE
Web page pre-fetching techniques are used to address the access latency problem of the Internet. In order to perform successful pre-fetching, a page ranking model is required pre-compute the next set of pages that are likely to be accessed by users. The PageRank algorithm is used to compute the importance of a set of Web pages based on their link structure, number of hit, time and area or location based. Some factors are adopted to personalize PageRank, so that it favors the pages that are more important to users. This research wok presents an improved PageRank algorithm that computes the PageRank values of the Web pages more precisely based on time and location based analysis.