Recent Developments in Artificial Intelligence and Communication Technologies

The traditional artificial guide service can be substituted by the advanced intelligent tourism guide system, which can help many developing tourism industries as the demand for tourism is going higher in today's world. An intelligent tourism guide system can create automatic recommendations according to the preferences [1]. With the instantaneous evolution of computer technology and electronic information technology as the basis, this chapter combines the tree-based algorithm and associated knowledge of tree theory to implement an algorithm and processing plan. The objective of our approach is to build a relationship between the user and the system. The application provides many services to the user meeting their needs and the purpose of gaining information about the places. The application mainly represents a mobile tour guide system with augmented reality. The main objective of the application is to make a system that runs on most of the mobile devices and becomes helpful to the user while visiting new places. The system should find a place using user preferences, like beaches, historical monuments, hill stations, temples, adventurous places, etc. The system should show recommendations about those places along with the description and images. This application will help the people who love to travel and want to travel to new places without having previous information about the place. This model [2-5] makes the use of efficient BST Searching as compared to the database. The information about various places is stored in the tree data structure, and it becomes easy to store a lot of data in the tree as compared to the database because it requires more memory and time to store lots of information into the database. The main advantage of using the system is to make the searching process easier and to ease the process of storing the data in the tree rather than the database. The tree-based algorithm is efficient in terms of storage and retrieval of data so that the performance of the system is enhanced. The application takes less time to fetch the data using a tree-based algorithm according to added preferences by the user as compared to the database, which takes more time to fetch the data and to display it as required.

With the tremendous increase in the use of internet in almost every sector of society, assigning addresses has appeared to be inefficient. The previous Internet Protocol version 4 (i.e. IPv4) failed to fulfill the highly growing demand. Though previous Internet Protocol Version 4 was used for assigning addresses, it could not sustain the high demand resulting in the downfall of IPv4. The chapter outlines the various advantages and disadvantages of the shift from IPv4 to IPv6. It also highlights the security threats of both the protocols and security issues due to which the coexistence of the two protocols was thought of as the solution. This chapter overall covers the transition from IPv4 to IPv6, their uses and management, followed by the analysis of IPv6 in terms of security issues. It further takes into account the effect of IPv6 on the world of Internet of Things (IoT). The eventual objective of this framework is to give comprehensive and detailed knowledge about the internet protocols in the Internet-of-Things world.

Recommendation systems are widely used today by online stores and various other leading sites, like Facebook, Instagram and LinkedIn, for providing suggestions to the users. The recommendation process helps the users to find the items that they may be interested in. Also, it is beneficial for the company to improve its overall profit. Recommendation engines use collaborative filtering technique or content-based approach to acquaint the users with such items. As these engines are so beneficial for users as well as for the trading websites, they have already been applied to a large number of fields, such as medical, education, tourism, finance, marketing and business; however, some areas are yet left unexplored. In this paper, we are presenting one such area where if recommendation engines are used, they can help a huge number of researchers around the globe. We propose a recommendation system that can help a number of scholars to get research papers based on the keyword entered by them, and the user will set a similarity index. This value of similarity will help in getting a limited number of papers from a huge pool.

Understanding the behavior of humans is a very important concern for social communication. Especially in real-time, predicting human activity and behavior has become the most vigorous research area in digital image processing and computer vision. To enhance the security in public and private domains in the field of humancomputer interaction and intelligent video surveillance, human behavior analysis is an important challenge in various applications. There are many basic approaches to analyze human activity, but recently, deep learning approaches have been shown that yield very interesting results in different domains. Human actions and behavior can be observed in the open as well as in sensitive areas, such as airports, banks, bus and train station, colleges, parking areas, etc., and prevent terrorism, theft, accidents, fighting, as well as other abnormal and suspicious activities through visual surveillance. This chapter thus seeks to reflect on methods of human activity recognition. This chapter presents a brief overview on human behavior recognition along with its challenges or issues and applications. Also, we have discussed the framework of recognition of suspicious human activity and various datasets used to train the system. The objective of this chapter is to provide general information about human behavior analysis and recent methods used in this field.

Over the last several decades, wireless sensor networks have grabbed a lot of attention because of their wide range of applications in scientific communities as well as industrial aspects. In WSN, sensor nodes are created with very limited resources, imposing energy constraints. Therefore, it is important to design a less energyconsuming, ascendible and power-efficient approach by selecting the optimal cluster heads (CHs) to enhance the life of these networks. Clustered sensor network is a method to optimize the power consumption in the network, which greatly affects the performance of the networks. In this article, we looked at clustering and routing issues by employing intelligent optimization techniques by considering the maximum attribute of the wireless sensor networks (WSN), which are conflicting in nature. The efficiency of WSN mainly depends upon the conflicting attributes, like residual energy, CH to base station distance, normal node to CH distance, etc. In this paper, multiattribute decision making (MADM) technique is considered for choosing the optimal CHs, so that energy consumption of the nodes is minimized and lifetime of the network can be maximized. The proposed approach is compared with other approaches like LEACH, HEED, etc. Results verified that the proposed algorithm is outmatched in comparison to existing algorithms.

Target tracking plays an important role in an application of WSN (Wireless Sensor Networks) with its growing popularity for various industrial as well commercial purposes. Many researches have been done for laying the framework of a target tracking algorithm to achieve accuracy while tracking down the target. Although the basic structure of a target tracking algorithm focuses on accuracy but the cooperation of both data accuracy and efficient energy utilization is quite distinct goal in the designing of target tracking algorithms. Through this survey, we present some recent target-based tracking algorithms that focus on dual goal, i.e., efficient power consumption along with the data accuracy. In this paper, we present a comparative structure that showcases that all surveyed algorithms are still on the verge of improved tracking accuracy with energy efficiency and there is still a need of optimization in terms of energy-efficient usage. The primary focus of the paper is to survey the level of energy efficiency in different categories of target tracking algorithms laid down in Wireless Sensor Networks.

Mobile learning technology is playing a vital role in today's education due to COVID-19. The ongoing pandemic has increased its demand and usage everywhere in the world. It is a technology that enables students to learn, interact, collaborate, and access education from any place and at any time by using internet-enabled mobile devices. These mobile devices include laptops, tablets, and mobile phones for distance learning. In the current scenario, this distance learning has proved as a boon, especially in the education and corporate sector widely. However, M-learning can also be applied in other fields, such as financial sectors, including banking, and non-financial sectors, including healthcare, industrial sector, etc. Mobile learning acceptance is a requirement that is still in progress. It has been adopted more in urban schools and higher education institutions in comparison to rural and secondary schools. Our study explores the evolution of M-learning, benefits, challenges, platforms, and various factors that affect mobile learning adoption. We have also discussed the existing models and frameworks for learning in higher education, university, etc. Hence, in this study, we have focused both on the positive and negative outcomes of all the previous studies to eliminate the issues and increase mobile learning adoption in the future. The main highlight of this paper is that we have presented various online platforms available for individuals, organizations, and enterprises for learning. These tools are as important as considering various other factors that have been discussed in previous findings. Moreover, a comparison table of various factors and conclusions of different researchers is shown. Furthermore, it has been concluded that the adoption of mobile learning technology must be increased in developing countries.

Soft computing has arisen as a standard presuming perspective; fuzzy ideas have advanced as imperative in the field of processing. The present chapter would expect to motivate and support the reader by giving all the necessary flavors empowering him to seek after and exude creative thoughts in the field of the fuzzy framework. The chapter presents the fundamental fuzzy ideas, including the operations performed on uncertainty sets. Fuzzy ideas and operations are contrasted with crispy sets for the better cognizance of the pursuers, specifically the naive. How the probabilistic and fuzzy frameworks (level of truthness) vary would be underscored with sufficient situations. Uncertainty, vagueness in the data, and historical crispy data when utilized later, would connect some vulnerability with it; an embodiment of such vulnerability and its need to consider in the processing is managed in detail. Different fuzzy membership functions commonly used are explored here. The underlying terminology (core, support and boundary) and its significance in the membership functions are explored. In the last segment of the section, text categorization (TC) application utilizing the fuzzy ideas in processing is investigated. The features in TC are transformed into fuzzy collections (soft, hard and blended) as feature reduction. Gaussian function is employed in the process of obtaining the fuzzy collections. Itemized steps during the time spent treating word-based data classification are talked about.

Highly demanding services require an appropriate amount of resources to manage the fluctuating workload in cloud environment, which is a challenging task for cloud service provides over the Internet. Cloud providers offer these services to enduser with pay and use model, such as utility computing. The services are offered to end-user by a cloud provider in a shareable fashion over Infrastructure-as-a-Service. So, IaaS is a type of computing service on which third parties host their application on virtualized platforms, such as either VMs or Containers. Whenever some containers are overloaded or under-loaded, it may cause SLA violation, degrade performance, cosume maximum energy, and also cause minimum throughput and maximum response time. It also leads to minimizing the customer satisfaction level along with cloud providers, leading to the penalty. The services hosted on VMs or Containers are highly demanding services, and these highly demanding services are handled with the help of load balancing. Load balancing is a way to automatically transfer the incoming requests or load across a group of back-end containers. It improves the distribution of workload across multiple virtual machines. Traditionally, load balancing algorithms use one or two parameters to balance the load. In this paper, we used one of the popular optimization techniques, namely the Technique for Order of Preferences by Similarity to Ideal Solution (TOPSIS) algorithm to manage the incoming traffic with the multiplecriteria decision-making (MCDM) technique. When the proposed technique was compared with different other techniques, such as round robin, it was found that TOPSIS gives better performance in terms of efficient resources utilization. It also minimizes the average response time, which prevents the machine from getting overloaded.

Information retrieval (IR) is to identify documents of non-uniform behavior that fulfill information requirements from the huge repository (maintained in computer systems). Different models have been defined to retrieve/fetch information. For example, the Boolean model, the Statistical model, which focuses on the vector space and probabilistic retrieval, and the Linguistic and Knowledge-based retrieval models. The Boolean model is defined as the “perfect match” model. If the queries are not accurate, they retrieve/fetch some irrelevant documents. This is called the precision (p) rate, which is the proportion of the relevant retrieved documents. The Boolean method provides good techniques to elaborate or concise a query. The Boolean method works well for the search process because of the clarity between the concepts. The Boolean retrieval model processes the queries in which terms of the queries are in the form of Boolean expressions, that is, in which terms of the user query combined with AND(&), OR(||), and NOT(!) operators. The model views documents in the form of inverted indexes. The key concept of an inverted index is to maintain a dictionary of terms. For every term, there is a collection of documents in which the term occurs. Posting is a collection of documents in which a term occurs. The list is known as the postings list (or inverted list), and all the postings lists are collectively called postings.