ABSTRACT Blockchain and artificial intelligence (AI) are transforming the financial industry by offering novel solutions to long‐standing challenges such as security risks, regulatory compliance, and operational inefficiencies. Despite significant progress, a systematic understanding of the intersection between blockchain and AI in financial risk mitigation is lacking. This study conducts a systematic literature review (SLR) to identify current applications, emerging trends, and security concerns at this intersection. Guided by four research questions, the review examines 134 peer‐reviewed publications sourced from IEEE Xplore, Springer, ScienceDirect, and Wiley databases. A PRISMA framework was used to ensure transparency in article selection. The analysis highlights key areas such as fraud detection, smart contract automation, decentralized identity verification, and regulatory challenges. Our findings contribute to the development of secure, AI‐enhanced blockchain solutions tailored to the financial sector. This review offers practical implications for policymakers, researchers, and financial institutions aiming to foster trust, resilience, and innovation in digital finance. The review's findings underline that the convergence of AI and blockchain offers both opportunities and emerging risks for financial institutions. Unlike previous reviews that examined these technologies separately, this study provides a consolidated understanding of how AI enhances blockchain's resilience and regulatory compliance in digital finance. The results also identify open challenges that should guide future empirical research, including model explainability, governance automation, and cross‐jurisdictional regulation.

ABSTRACT Effective management of software assets in their whole lifespan is the main goal of software asset management (SAM) and is a contemporary organizational practice. It includes a range of tasks such as purchasing, implementing, and maintaining software inside a company. SAM seeks to minimize the risks and expenses related to software ownership while ensuring that software resources are used as efficiently as possible to support business activities. Recently, a great increase in the use of this technique has occurred, especially in large‐scale enterprises where the complexity and diversity of software assets have faced major hurdles. The proposed study presents an overview of the analysis of the recent approaches and hurdles in the area of SAM. Because big software companies have access to a multitude of resources and experience, maximizing the reuse of software assets inside these organizations is a common topic of academic and industrial study. Through the integration of several important attributes from previous research endeavors, the current study seeks to determine the most common attributes for the research. The study aims to contribute to the area by integrating the Analytical Hierarchy Process (AHP) along with the Weighted Aggregated Sum Product Assessment (WASPAS) approaches to give a rigorous and systematic way to analyze and rate the prominent qualities for selection of the most appropriate choice among the available alternatives.

Abstract The advancement in the field of artificial intelligence (AI) has revolutionized every field of life, including the learning of second languages. These intelligent devices are capable of effectively and efficiently utilizing the time and energy of both learners and teachers. Students can learn at their own pace and at their own skill level. They can learn and practice in an interactive and fruitful environment thanks to intelligent chatbots and voice assistants. Students no longer require humanized teachers as a result of the use of these new methodologies; instead, they can learn more effectively by interacting with computer‐assisted systems. With the integration of information and communication technology (ICT) and AI, new technologies like computer‐assisted language learning (CALL) and mobile‐assisted language learning (MALL) are playing a very crucial role in the learning of the English language. Due to the various AI‐based applications and technologies available, learners are unable to use the most valuable and effective ones. This paper focuses on the role of AI in the learning of the English language. The study will help learners in the selection of efficient and effective AI‐powered paradigms for the teaching and learning process of the English language. Various features have been selected from the identified ones, and then, on the basis of these features, different AI‐grounded paradigms for English learning are ranked using analytical hierarchy process (AHP) and Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS). The alternative with the highest performance value is ranked at the top of all available alternatives, while the one with the lowest performance score is placed at the last.

Abstract Cloud computing is characterized as a shared computing and communication infrastructure. It encourages the efficient and effective developmental processes that are carried out in various organizations. Cloud computing offers both possibilities and solutions of problems for outsourcing and management of software developmental operations across distinct geography. Cloud computing is adopted by organizations and application developers for developing quality software. The cloud has the significant impact on utilizing the artificial complexity required in developing and designing quality software. Software developmental organization prefers cloud computing for outsourcing tasks because of its available and scalable nature. Cloud computing is the ideal choice utilized for development modern software as they have provided a completely new way of developing real‐time cost‐effective, efficient, and quality software. Tenants (providers, developers, and consumers) are provided with platforms, software services, and infrastructure based on pay per use phenomenon. Cloud‐based software services are becoming increasingly popular, as observed by their widespread use. Cloud computing approach has drawn the interest of researchers and business because of its ability to provide a flexible and resourceful platform for development and deployment. To determine a cohesive understanding of the analyzed problems and solutions to improve the quality of software, the existing literature resources on cloud‐based software development should be analyzed and synthesized systematically. Keyword strings were formulated for analyzing relevant research articles from journals, book chapters, and conference papers. The research articles published in (2011–2021) various scientific databases were extracted and analyzed for retrieval of relevant research articles. A total of 97 research publications are examined in this SLR and are evaluated to be appropriate studies in explaining and discussing the proposed topic. The major emphasis of the presented systematic literature review (SLR) is to identify the participating entities of cloud‐based software development, challenges associated with adopting cloud for software developmental processes, and its significance to software industries and developers. This SLR will assist organizations, designers, and developers to develop and deploy user‐friendly, efficient, effective, and real time software applications.

Abstract Art design is a method of conveying human feelings and emotions, particularly via the use of visual structures such as paintings or sketches. Every element of our life, including arts and crafts, has been positively affected by the introduction of novel technologies such as artificial intelligence (AI) and machine learning (ML). In today's modern environment, these technologies have altered the techniques of art creation, consumption, and distribution. In today's environment, ML and human emotions are the two most important aspects for interactive and high‐quality art design. Whereas traditional learning systems can be extremely effective in the teaching and learning process of art‐related subjects, AI and ML can be very effective in the teaching and learning process of art‐related subjects for the advancement of learners' artistic skills. The productive and active role of AI and ML in the developments and advancements of art design has been given a very comprehensive and detailed overview in this study. Following a detailed examination of the existing techniques, distinct characteristics have been identified. Six of the most widely utilized features were chosen from the literature to execute the analytical hierarchy process (AHP). For ranking the options based on the weights derived by AHP, the TOPSIS algorithm is used. The option with the best performance came in first, whereas the one with the worst performance came in last.

The future communication technologies like 6G are capable to provide higher mobility and better quality-of-service requirements to Internet of Things (IoT). To ensure mobility, the 6G technologies need more reliable and scalable solutions, which are capable to integrate large-scale heterogeneous IoT networks. In a heterogeneous environment, seamless mobility along with the demands of IP addresses requires a proxy mobile IPv6 (PMIPv6) protocol that provides cost-effective solutions in next-generation IoT networks. The PMIPv6 has been exploited for resource efficiency in IoT-enabled next-generation networks. In this article, we have proposed a demand-based resource-efficient location-aware PMIPv6 extension for seamless mobility in the next-generation IoT networks. The proposed approach efficiently utilizes the network resources using location information and received signal strength (RSS). This solution enhances the performance of the PMIPv6 protocol in terms of signaling cost, and load on network entities. Furthermore, mathematical models are derived in terms of signaling cost and load distribution. The proposed solution is compared with the existing RSS-based PMIPv6 extension protocols. The results show that the proposed scheme enhances the performance and is a resource-friendly for the next-generation large-scale IoT networks.

Regional and cultural diversities around the world have given birth to a large number of writing systems and scripts, which consist of varying character sets. Developing an optimal character recognition for such a varying and large character set is a challenging task. Unlimited variations in handwritten text due to mood swings, varying writing styles, changes in medium of writing, and many more puzzle the research community. To overcome this problem, researchers have proposed various techniques for the automatic recognition of cursive languages like Urdu, Pashto, and Arabic. With the passage of time, the field of text recognition matured, and the number of publications exponentially increased in the targeted field. It is very difficult to find all the techniques developed, calculate the time and resource consumptions, and understand the cost–benefit tradeoffs among these techniques. These tradeoffs resist making this technology able for practical use. To address these tradeoffs, this article systematic analysis to identify gaps in the literature and suggest new enhanced solution accordingly. A total of 153 of the most relevant articles from 2008 to 2022 are analyzed in thissystematic literature review (SLR)work. This systematic review process shows (1) the list of techniques suggested for cursive text recognition purposes and its capabilities, (2) set of feature extraction techniques proposed, and (3) implementation tools used to design and simulate the empirical studies in this specialized field. We have also discussed the emerging trends and described their implications for the research community in this specialized domain. This systematic assessment will ultimately help researchers to perform an overview of the existing character/text recognition approaches, recognition capabilities, and time consumption and subsequently identify the areas that requires a significant attention in the near future.

The characteristic of context dependency in Urdu words needs to be handled carefully while performing Urdu sentiment analysis. In this research, an already constructed Urdu sentiment lexicon of positive and negative words is further expanded by the addition of context-dependent words. These context-dependent words are used with or without conjunctions. Rules are formulated for assigning polarities to those context-dependent words that are surrounded by the positive or negative words. These rules were incorporated in the Urdu sentiment analyzer. Fusion of these rules for handling context-dependent words and the expanded Urdu sentiment lexicon resulted in increasing the accuracy of the Urdu sentiment analyzer from 83.43% to 89.03% with 0.8655 precision, 0.9053 recall, and 0.8799 F-measure, which is a statistically significant improvement.

Billions of devices are connected via the Internet which has produced various challenges and opportunities. The increase in the number of devices connected to the Internet of things (IoT) is nearly beyond imagination. These devices are communicating with each other and facilitating human life. The connection of these devices has provided opening directions for the smart applications which are one of the growing areas of research. Among these opportunities, security and privacy are considered to be one of the major issues for researchers to tackle. Proper security measures can prevent attackers from interrupting the security of IoT network inside the smart city for secure data traffic. Keeping in view the security consideration of data traffic for smart devices and IoT, the proposed study presented machine learning algorithms for securing the data traffic based on a firewall for smart devices and IoT network. The study has used the dataset of “Firewall” for validation purposes. The experimental results of the approach show that the hybrid deep learning model (based on convolution neural network and support vector machine) outperforms than decision1 rules and random forest by generating a recognition rate of 95.5% for the hybrid model, 68.5% for decision rules, and 78.3% accuracy for random forest. The validity of the proposed model is also tested based on other performance metrics such as f score, error rate, recall, and precision. This high accuracy rate and other performance values show the applicability of the proposed hybrid model to secure data traffic purposes in smart devices. This can be used in many research areas of the smart city for security purposes.

In computer vision and artificial intelligence, text recognition and analysis based on images play a key role in the text retrieving process. Enabling a machine learning technique to recognize handwritten characters of a specific language requires a standard dataset. Acceptable handwritten character datasets are available in many languages including English, Arabic, and many more. However, the lack of datasets for handwritten Pashto characters hinders the application of a suitable machine learning algorithm for recognizing useful insights. In order to address this issue, this study presents the first handwritten Pashto characters image dataset (HPCID) for the scientific research work. This dataset consists of fourteen thousand, seven hundred, and eighty-four samples—336 samples for each of the 44 characters in the Pashto character dataset. Such samples of handwritten characters are collected on an A4-sized paper from different students of Pashto Department in University of Peshawar, Khyber Pakhtunkhwa, Pakistan. On total, 336 students and faculty members contributed in developing the proposed database accumulation phase. This dataset contains multisize, multifont, and multistyle characters and of varying structures.

Software component plays a significant role in the functionality of software systems. Component of software is the existing and reusable parts of a software system that is formerly debugged, confirmed, and practiced. The use of such components in a newly developed software system can save effort, time, and many resources. Due to the practice of using components for new developments, security is one of the major concerns for researchers to tackle. Security of software components can save the software from the harm of illegal access and damages of its contents. Several existing approaches are available to solve the issues of security of components from different perspectives in general while security evaluation is specific. A detailed report of the existing approaches and techniques used for security purposes is needed for the researchers to know about the approaches. In order to tackle this issue, the current research presents a systematic literature review (SLR) of the present approaches used for assessing the security of software components in the literature by practitioners to protect software systems for the Internet of Things (IoT). The study searches the literature in the popular and well-known libraries, filters the relevant literature, organizes the filter papers, and extracts derivations from the selected studies based on different perspectives. The proposed study will benefit practitioners and researchers in support of the report and devise novel algorithms, techniques, and solutions for effective evaluation of the security of software components.

The interest in Facial Expression Recognition (FER) is increasing day by day due to its practical and potential applications, such as human physiological interaction diagnosis and mental disease detection. This area has received much attention from the research community in recent years and achieved remarkable results; however, a significant improvement is required in spatial problems. This research work presents a novel framework and proposes an effective and robust solution for FER under an unconstrained environment; it also helps us to classify facial images in the client/server model along with preserving privacy. There are a lot of cryptography techniques available but they are computationally expensive; on the other side, we have implemented a lightweight method capable of ensuring secure communication with the help of randomization. Initially, we perform preprocessing techniques to encounter the unconstrained environment. Face detection is performed for the removal of excessive background and it detects the face in the real-world environment. Data augmentation is for the insufficient data regime. A dual-enhanced capsule network is used to handle the spatial problem. The traditional capsule networks are unable to sufficiently extract the features, as the distance varies greatly between facial features. Therefore, the proposed network is capable of spatial transformation due to the action unit aware mechanism and thus forwards the most desiring features for dynamic routing between capsules. The squashing function is used for classification purposes. Simple classification is performed through a single party, whereas we also implemented the client/server model with privacy measurements. Both parties do not trust each other, as they do not know the input of each other. We have elaborated that the effectiveness of our method remains unchanged by preserving privacy by validating the results on four popular and versatile databases that outperform all the homomorphic cryptographic techniques.

Abstract Internet of Health Things (IoHT) is an extension of the Internet of Things (IoT), which plays an important role in the observation, consultation, monitoring, and treatment process of remote exchange data processes in healthcare. The integration, computation, and interoperability are facilitated through various sensors, controllers, and actuators. Security is considered as one of the important factors for the communication of these devices to smoothly run the activities of healthcare. The current literature associated to the security of IoHT cover diverse aspects, though there is dire need of the knowledge, which can thoroughly show the review of current state‐of‐the‐art work in a systematic way. To overcome this limitation, the planned study offerings a systematic literature review (SLR) for the IoHT security and provenance. This review will support researchers to take benefit from the present literature and devise novel solutions considering the current study as evidence in IoHT. This paper focused on the security analysis and provenance for the IoHT and answering the defined questions in the current study.

Abstract Crowdsourcing is simply the outsourcing of different tasks or work to a diverse group of individuals in an open call for the purpose of utilizing human intelligence. Crowdsourcing nowadays used to support and enhance software engineering in different aspects. In this proposed study, a systematic literature review was conducted for the last 10 years from 2010 to 2019. During the filtering process, a total of 120 relevant studies have been identified, and then the most relevant 70 studies were selected to include as part of the current study. The proposed study shows the effect of task assignment in crowdsourcing, such as if a task is assigned to an appropriate worker or an inappropriate worker, what will be the consequences. The study also highlights crowdsourcing usage in the field of software engineering. All the existing task assignment methods used for assigning the task to make crowdsourcing activity more effective have been analyzed. The study also highlights all the available crowdsourcing platforms used for a variety of task to be performed. The study concludes by identifying the issues regarding the task assignment and to specify the methods for enhancement in the assignment of tasks for future research in crowdsourcing.

Computation and data sensitivity are the metrics of the current Internet of Things (IoT). In cloud data centers, current analytics are often hosted and reported on suffering from high congestion, limited bandwidth, and security mechanisms. Various platforms are developed in the area of fog computing and thus implemented and assessed to run analytics on multiple devices, including IoT devices, in a distributed way. Fog computing advances the paradigm of cloud computing on the network edge, introducing a number of options and facilities. Fog computing enhances the processing, verdicts, and interventions to occur through IoT devices and spreads only the necessary details. The ideas of fog computing based on IoT in healthcare frameworks are exploited by shaping the disseminated delegate layer of insight between sensor hubs and the cloud. The cloud proposed a system adapted to overcome various challenges in omnipresent medical services frameworks, such as portability, energy efficiency, adaptability, and unwavering quality issues, by accepting the right to take care of certain weights of the sensor network and a distant medical service group. An overview of e‐health monitoring system in the context of testing and quality assurance of fog computing is presented in this paper. Relevant papers were analyzed in a comprehensive way for the identification of relevant information. The study has compiled contributions of the existing methodologies, methods, and approaches in fog computing e‐healthcare.

In this article, an efficient numerical approach for the solution of second‐order delay differential equations to deal with the experimentation of the Internet of Industrial Things (IIoT) is presented. With the help of the Haar wavelet technique, the considered problem is transformed into a system of algebraic equations which is then solved for the required results by using Gauss elimination algorithm. Some numerical examples for convergence of the proposed technique are taken from the literature. Maximum absolute and root mean square errors are calculated for various collocation points. The results show that the Haar wavelet method is an effective method for solving delay differential equations of second order. The convergence rate is also measured for various collocation points, which is almost equal to 2.

Context. The improvements made in the last couple of decades in the requirements engineering (RE) processes and methods have witnessed a rapid rise in effectively using diverse machine learning (ML) techniques to resolve several multifaceted RE issues. One such challenging issue is the effective identification and classification of the software requirements on Stack Overflow (SO) for building quality systems. The appropriateness of ML-based techniques to tackle this issue has revealed quite substantial results, much effective than those produced by the usual available natural language processing (NLP) techniques. Nonetheless, a complete, systematic, and detailed comprehension of these ML based techniques is considerably scarce. Objective. To identify or recognize and classify the kinds of ML algorithms used for software requirements identification primarily on SO. Method. This paper reports a systematic literature review (SLR) collecting empirical evidence published up to May 2020. Results. This SLR study found 2,484 published papers related to RE and SO. The data extraction process of the SLR showed that (1) Latent Dirichlet Allocation (LDA) topic modeling is among the widely used ML algorithm in the selected studies and (2) precision and recall are amongst the most commonly utilized evaluation methods for measuring the performance of these ML algorithms. Conclusion. Our SLR study revealed that while ML algorithms have phenomenal capabilities of identifying the software requirements on SO, they still are confronted with various open problems/issues that will eventually limit their practical applications and performances. Our SLR study calls for the need of close collaboration venture between the RE and ML communities/researchers to handle the open issues confronted in the development of some real world machine learning-based quality systems.

In this paper, we proposed LCX-MAC (local coordination X-MAC) as an extension of X-MAC. X-MAC is an asynchronous duty cycle medium access control (MAC) protocol. X-MAC used one important technique of short preamble which is to allow sender nodes to quickly send their actual data when the corresponding receivers wake up. X-MAC node keeps sending short preamble to wake up its receiver node, which causes energy, increases transmission delay, and makes the channel busy since a lot of short preambles are discarded, as these days Internet of Things (IoT) healthcare with different sensor nodes for the healthcare is time-critical applications and needs a quick response. A possible improvement over X-MAC is that local information of each node will share with its neighbour node. This local information exchanged will cause much less overhead than in the nodes which are synchronized. To calculate the effect of this the local coordination on X-MAC in this paper, we built an analytical model of LCX-MAC that incorporates the local coordination in X-MAC. The analytical results show that LCX-MAC outperformed X-MAC and X-MAC/BEB in terms of throughput, delay, and energy.

With the development of communication systems, information securities remain one of the main concerns for the last few years. The smart devices are connected to communicate, process, compute, and monitor diverse real-time scenarios. Intruders are trying to attack the network and capture the organization’s important information for its own benefits. Intrusion detection is a way of identifying security violations and examining unwanted occurrences in a computer network. Building an accurate and effective identification system for intrusion detection or malicious activities can secure the existing system for smooth and secure end-to-end communication. In the proposed research work, a deep learning-based approach is followed for the accurate intrusion detection purposes to ensure the high security of the network. A convolution neural network based approach is followed for the feature classification and malicious data identification purposes. In the end, comparative results are generated after evaluating the performance of the proposed algorithm to other rival algorithms in the proposed field. These comparative algorithms were FGSM, JSMA, C&W, and ENM. After evaluating the performance of these algorithms and the proposed algorithm based on different threshold values ranging, Lp norms, and different parametric values for c, it was concluded that the proposed algorithm outperforms with small Lp values and high Kitsune scores. These results reflect that the proposed research is promising toward the identification of attack on data packets, and it also reflects the applicability of the proposed algorithms in the network security field.

The spam detection is a big issue in mobile message communication due to which mobile message communication is insecure. In order to tackle this problem, an accurate and precise method is needed to detect the spam in mobile message communication. We proposed the applications of the machine learning-based spam detection method for accurate detection. In this technique, machine learning classifiers such as Logistic regression (LR), K-nearest neighbor (K-NN), and decision tree (DT) are used for classification of ham and spam messages in mobile device communication. The SMS spam collection data set is used for testing the method. The dataset is split into two categories for training and testing the research. The results of the experiments demonstrated that the classification performance of LR is high as compared with K-NN and DT, and the LR achieved a high accuracy of 99%. Additionally, the proposed method performance is good as compared with the existing state-of-the-art methods.

Internet of Medical Things (IoMT) plays an important role in healthcare. Different devices such as smart sensors, wearable devices, handheld, and many other devices are connected in a network in the form of Internet of Things (IoT) for the smooth running of communication in healthcare. Security of these devices in healthcare is important due to its nature of functionality and efficiency. An efficient and robust security system is in dire need to cope with the attacks, threats, and vulnerability. The security evaluation of IoMT is an issue since couple of years. Therefore, the aim of the proposed study is to evaluate the security of IoMT by using the analytic network (ANP) process. The proposed approach is applied using ISO/IEC 27002 (ISO 27002) standard and some other important features from the literature. The results of the proposed research demonstrate the effective IoMT components which can further be used as secure IoMT.

Feature-based software birthmark is an essential property of software that can be used for the detection of software theft and many other purposes like to assess the security in end-to-end communication systems. Research on feature-based software birthmark shows that using the feature-based software birthmark joint with the practice of software birthmark estimation together can deliver a right and influential method for detecting software piracy and the amount of piracy done by a software. This can also guide developers in improving security of end-to-end communication system. Modern day software industry and systems are in demand to have an unbiased method for comparing the features-based birthmark of software competently, and more concretely for the detecting software piracy and assessing the security of end-to-end communication systems. In this paper, we proposed a mathematical model, which is based on a differential system, to present feature-based software birthmark. The model presented in this paper provides an exclusive way for the features-based birthmark of software and then can be used for comparing birthmark and assessing security of end-to-end communication systems. The results of this method show that the proposed model is efficient in terms of effectiveness and correctness for the features-based software birthmark comparison and security assessment purposes.

Components are the significant part of a system which plays an important role in the functionality of the system. Components are the reusable part of a system which are already tested, debugged, and experienced based on the previous practices. A new system is developed based on the reusable components, as reusability of components is recommended to save time, effort, and resources as such components are already made. Security of components is a significant constituent of the system to maintain the existence of the component as well as the system to function smoothly. Component security can protect a component from illegal access and changing its contents. Considering the developments in information security, protecting the components becomes a fundamental issue. In order to tackle such issues, a comprehensive study report is needed which can help practitioners to protect their system. The current study is an endeavor to report some of the existing studies regarding component security evaluation based on multicriteria decision and machine learning algorithms in the popular searching libraries.

The impact of Internet of Things has been revolutionized in all fields of life, but its impact on the healthcare system has been significant due to its cutting edge transition. The role of Internet of Things becomes more dominant when it is supported by the features of mobile computing. The mobile computing extends the functionality of IoT in healthcare environment by bringing a massive support in the form of mobile health (m-health). In this research, a systematic literature review protocol is proposed to study how mobile computing assists IoT applications in healthcare, contributes to the current and future research work of IoT in the healthcare system, brings privacy and security in health IoT devices, and affects the IoT in the healthcare system. Furthermore, the intentions of the paper are to study the impacts of mobile computing on IoT in healthcare environment or smart hospitals in light of our systematic literature review protocol. The proposed study reports the papers that were included based on filtering process by title, abstract, and contents, and a total of 116 primary studies were included to support the proposed research. These papers were then analysed for research questions defined for the proposed study.

Abstract Just like other languages, a large number of intensifiers are used in Urdu language. There may be a single occurrence of an intensifier or there may be multiple consecutive occurrences. While performing sentiment analysis of Urdu text, these intensifiers need special treatment for obtaining more accurate results, which is the main focus of this research work. A wide coverage Urdu sentiment lexicon is developed where intensifiers are identified and placed in a separate file. While developing Urdu sentiment analyser, rules are specifically formulated for assigning polarities to the intensifiers in text, if they are surrounded by the positive or negative words. Results show that the method proved to be effective in attaining the correct classification of Urdu sentences as positive, negative, or neutral, compared with traditional methods. Implementation of rules for intensifiers increased the accuracy of Urdu sentiment analyser from 78.33% to 83.42%, which is a statistically significant improvement. It is concluded that intensifiers cannot be ignored while performing sentiment analysis. Effective handling of intensifiers can significantly improve the performance of sentiment analyser.