Customer Support System Research Articles

Optimizing hierarchical classifiers with parameter tuning and confidence scoring

Hierarchical classifiers play a crucial role in addressing complex classification tasks by breaking them down into smaller, more manageable sub-tasks. This paper continues a series of works, focused on the technical Ukrainian texts hierarchical classification, specifically the classification of repair works and spare parts used in automobile maintenance and servicing. We tackle the challenges posed by multilingual data inputs – specifically Ukrainian, Russian, and their hybrid – and the lack of standard data cleaning models for the Ukrainian language. We developed a novel classification algorithm, which employs TF-IDF victimization with unigrams and bigrams, keyword selection, and cosine similarity for classification. This paper describes a method for training and evaluating a hierarchical classification model using parameter tuning for each node in a tree structure. The training process involves initializing weights for tokens in the class tree nodes and input strings, followed by iterative parameter tuning to optimize classification accuracy. Initial weights are assigned based on predefined rules, and the iterative process adjusts these weights to achieve optimal performance. The paper also addresses the challenge of interpreting multiple confidence scores from the classification process, proposing a machine learning approach using Scikit-learn's GradientBoostingClassifier to calculate a unified confidence score. This score helps assess the classification reliability, particularly for unlabeled data, by transforming input values, generating polynomial parameters, and using logarithmic transformations and scaling. The classifier is fine-tuned using hyper parameter optimization techniques, and the final model provides a robust confidence score for classification tasks, enabling the verification and classification results optimization across large datasets. Our experimental results demonstrate significant improvements in classification performance. Overall classification accuracy nearly doubled after training, reaching 92.38 %. This research not only advances the theoretical framework of hierarchical classifiers but also provides practical solutions for processing large-scale, unlabeled datasets in the automotive industry. The developed methodology can enhance various applications, including automated customer support systems, predictive maintenance, and decision-making processes for stakeholders like insurance companies and service centers. Future work will extend this approach to more complex tasks, such as extracting and classifying information from extensive text sources like telephone call transcriptions.

Leveraging Intent Detection and Generative AI for Enhanced Customer Support

Customer support plays a pivotal role in shaping customer satisfaction and fostering loyalty within any business. This paper delves into how the integration of intent detection and generative AI (GenAI) can transform customer support systems. At the core of this transformation is the ability to understand user intent, which is essential for directing customers effectively through the support funnel to the appropriate services. By employing sophisticated natural language processing (NLP) techniques, training LLM to perform RAG and machine learning models, businesses can precisely discern customer intents. This capability allows for the delivery of tailored, immediate responses. The paper further explores the methodologies employed, the advantages gained, and the challenges faced in the adoption of these advanced technologies in customer support systems.

Conversational AI: An Explication of Few-Shot Learning Problem in Transformers-Based Chatbot Systems

With the recent advancements in conversational artificial intelligence (AI), the practical applications of chatbots have risen significantly in diverse domains such as healthcare, education, e-government, customer support systems, social platforms, and entertainment. The chatbot converses with humans in natural language and responds to their queries with precise and relevant answers. The relevant literature presents several methods for intent classification and slot mapping for chatbot question answering. However, the existing chatbot architectures still suffer from few-shot learning problem (imbalance ratio of class labels over samples) and ineffective dialog management to retain the context and slots mapping. This study aims to introduce the architecture of chatbot by focusing few-shot learning problem and context management in dialog-based conversations. First, to mitigate the few-shot learning problem with chatbots, we propose a novel hybrid intent and slots transformers (HIST) model. The HIST chatbot architecture utilizes transformers and self-attention mechanisms along with bigated recurrent unit and combines conditional random field algorithms for intent classification and slot extraction. Second, to address dialog management, we introduce a hybrid interaction strategy for slots mapping and effective conversational context management. To validate the proposed model’s effectiveness, comprehensive empirical analysis is carried out using three benchmark datasets including airline travel information system (ATIS), banking77, and conversational language interface for natural conversation 150 (CLINC150). The results show that HIST outperforms against the state-of-the-art existing methods with a clear margin and obtained an accuracy of 94.89% and 96.17% for intent classification and slots extraction, respectively. The empirical results confirm the effectiveness of the HIST chatbot for resolving the few-shot learning problem with effective dialog management in chatbot systems.

ASSESSING THE VALUE OF BLACK FRIDAY PROMOTIONS: AN ANALYSIS OF INSTAGRAM USERS’ SENTIMENTS AND BEHAVIORAL RESPONSES

Black Friday, as a significant global retail phenomenon, provides substantial insights into consumer behavior and the effectiveness of marketing strategies. This study explores into the dynamics of consumer engagement by analyzing user-generated content (UGC) on Instagram, focusing on the 2021 Black Friday promotions by key technology companies in Turkey. Utilizing an advanced three-step text mining methodology, the research commences with Latent Dirichlet Allocation (LDA) for organizing data into distinct thematic clusters pertinent to Black Friday promotions. This is followed by a sentiment analysis, executed using Python, to evaluate the emotional nuances of the UGC in relation to these themes and the corresponding company promotions. The concluding phase involves an exhaustive textual analysis (TA) to extract actionable insights, which are instrumental in refining promotional strategies and deepening the comprehension of consumer interactions on social media platforms. The results reveal a predominantly positive reception of exclusive promotions and smartphone deals, highlighting their effectiveness as strategic elements in social media marketing. In contrast, themes linked to perceived fraud, negative feedback, misinformation, and customer service issues elicited adverse reactions from consumers. These contrasting responses emphasize the imperative for brands to develop transparent, authentic marketing communications and robust customer support systems. The study not only offers strategic recommendations for brands aiming to enhance their social media campaigns but also contributes a theoretical framework for future research in digital consumer behavior, especially in the context of significant promotional events like Black Friday.

ПЕРСПЕКТИВИ РОЗВИТКУ FINTECH В УКРАЇНІ ТА СВІТІ У 2024 РОЦІ

The main trends in the development of FinTech in Ukraine and the in world in 2024 are analyzed in the article. Three areas in the FinTech development are currently identified as the most relevant: use of artificial intelligence, the latest technologies, and changes in financial services. Each area is analyzed and characterized in terms of content and development prospects. The impact of the regulator, the National Bank of Ukraine, on the activity of FinTech companies is considered separately. Plans to attract investments by Ukrainian FinTech companies in 2024 are analyzed: 45% of companies plan to develop by their own expense, and 39% plan to attract private investors’ resources. Plans for Ukrainian FinTech companies to enter the international market this year were analyzed: almost half of Ukrainian FinTech companies (47%) are already operating in international markets, and 38% plan to enter in the future. The analysis of the dynamics of the global artificial intelligence market in 2021-2023 and its forecast until 2030 shows that in 2023 the global artificial intelligence market was evaluated at about USD 200 billion. According to the Statistics Portal, this market is expected to grow significantly to almost $2 trillion by 2030. Based on the analysis of literature sources, it has been determined that new technologies that become popular in the FinTech sector will be: open banking, e-wallets, digital currencies, etc. To support their systemic development, it is necessary to use digital technologies. Thus, the article identifies promising areas of artificial intelligance use in the process of FinTech development in banking institutions: automated trading and investments (in the short term, investors will be able to focus on the development and implementation of specific strategies by outsourcing trading technologies to artificial intelligence), customer support system (text, voice, and video chatbots will be improved through integration with ChatGPT), fraud protection (automatic analysis of large amounts of data to identify potentially dangerous transactions will take much less time than manual analysis).

Investigating the influence of omnichannel retailing on consumer decision making in the Jordanian market

The present research investigates the use of omnichannel retailing and its influence on the Jordanian market, focusing on four important dimensions: integration, fulfilment, usability, and seamlessness. This study attempts to answer how strategies of omnichannel affect purchase-decisions and satisfaction of customers among students at private universities in Jordan as expectations for integrated purchase experiences that combine physical and digital channels continue to rise. Using Structural Equation Modelling (SEM), this study empirically examines the relationships between various dimensions of omnichannel retailing and consumer decision-making. The findings reveal that strong integration across sales channels, fast delivery options, simple interfaces, and seamless transitions significantly enhance consumer decision-making. This underscores the importance of robust infrastructure facilities, dependable information availability options, flexible fulfillment methods, user-friendly interface designs as well as a smooth customer support system which all together make for a great shopping experience. Thus, local retailers are given clear recommendations to invest in current systems by diversifying their delivery alternatives, improving mobile platforms, and using data analytics to match marketing efforts. The main focus of this paper is effective merchandising practices that can help retailers connect with consumers effectively during changing market situations. Moreover, there are several lessons from this research for businesses operating under stiff competition but with the increasing needs of modern-day consumers.

Implementasi Sistem Layanan Dukungan dan Pengaduan Pelanggan pada Produk Justmine Beauty Berbasis Mobile

Justmine Beauty is a series of daily skin care products that can overcome all skin types, natural ingredients taken from water can provide nutrition and nourish facial skin, then specially formulated and dermatologically tested which has halal certification and BPOM so that it is suitable for skin problems and accompany activities, keep facial skin healty with Justmine Beauty. Generally, customers conduct consultations only using WhatsApp or other social media where later this consultation history is also lost because it is not saved. Customer complaints contained in the brans mostly contain complaints of damageto the product. The author’s foal for making this research reportis to implement a customer support and complaint service system for Justmine Beauty products. Based on the result of the application that has been built, researchers can display consultation and complaint services so that users can consult and make mobile-based complaints.

Automation of Customer Support System (Chatbot) to Solve Web Based Financial and Payment Application Service

One of the most important features of any online service is the quality of its customer care. However, with the development of NLP tools, businesses are considering automated chatbot solutions to keep up with the increasing demand for their products and services. In view of this, the chatbot was developed using AIML java interpreter library Program AB which helps match input and output predefined in the AIML file. AIML (Artificial Intelligence Markup Language) was used to preprocess and train the bot using ready-made AIML file for FAQ questions. Also, vaadin was used to build a web UI to interact with the trained AIML bot. Finally, a google script was written to translate from any language to English for the bot to understand and send the response in the preferred language of the user. Findings showed that the response time of the bot is dependent of the network, as the design gave a score of 70%, 80%, 90% and 90% for load testing, stability, reliability testing and usability testing, respectively. Also, the bot is compatible with different operating systems, both for forward compatibility and backward compatibility having a score of 95%. The bot was able to answer customer questions, enquiries and complaints and the response time of the bot depends on the strength of the network since it is web based. Hence, the system provided a simple, cheaper, and durable customer financial and payment application service. Since chatbots cannot answer all questions, businesses that decide to implement them should ensure that they have enough protections in place against attacks and that routine requests are standardised to ensure optimal performance.

Centralized End User Support System using Rapid API and Firebase with BOT Framework

A Centralized customer support system is a software platform designed to manage and streamline customer interactions with a company. This system is typically used by customer service teams to provide support, handle inquiries, and resolve issues. The main objective of a centralized customer support system is to improve the overall customer experience and satisfaction by enabling prompt and effective communication between customers and the company. The system can include various features, such as a knowledge base, automated responses, live chat, and ticket management, all aimed at ensuring that customers receive the best possible service. A centralized customer support system is an essential tool for any business looking to enhance their customer service capabilities and build long-term relationships with their clients.

A multi-criteria decision analysis model for selecting an optimum customer service chatbot under uncertainty

Chatbots are increasingly integrated into customer support systems due to recent developments in artificial intelligence. Chatbots can provide quick and tailored services to users, including employees and consumers. This paper presents a decision-making framework for chatbot selection in the telecommunication industry. The most suitable chatbot is chosen using a new combination strategy combining Combined Compromise Solution (CoCoSo) and analytic hierarchy process (AHP) in the context of single-valued neutrosophic sets (SVNSs). The proposed method uses the AHP to generate attribute weights (collected from previous studies), and CoCoSo ranks the options and selects the best chatbot. The AHP-CoCoSo is an efficient multi-criteria decision-making model for solving complex under uncertainty. The usefulness and viability of the proposed model are statistically demonstrated by solving an explanatory case study of chatbot selection in telecommunication. The method’s resilience and strength are further demonstrated through sensitivity analysis and comparisons with previously proposed approaches. The findings of this study reveal that the approach used might lead to more realistic results when dealing with uncertain information.

Analyzing the Use of Artificial Intelligence and MachineLearning in Customer Support Systems

In order to identify areas that need further research, this study will analyze how machine learning (ML) and artificial intelligence (AI) technologies are applied in various customer services. This is done by using a systematic review of literature procedure to examine publications on the application of AI and ML in customer service that have been posted on various academic websites. Numerous AI and ML approaches may be applied by businesses to improve customer support and assistance. The use of self-service technologies, integrated product-service offers, service excellence, and word-of-mouth marketing are some of these techniques. The study’s conclusion addresses the customer service knowledge gaps that it identified when reviewing the literature. The information provided by the current study on how Artificial Intelligence (AI) and Machine Learning (ML) technologies may be applied or properly used to improve customer service or may be fruitful to businesses. The findings of the study will provide theoretical and practical implication of AI and ML in customer support system to develop more effective and efficient strategies to the users and policy makers. Ultimately the findings of the study will help increase the degree of satisfaction of customer. Longitudinal studies should assess AI and ML’s long-term impact on customer support systems, balancing human-AI collaboration, developing transparent models, considering cross-cultural factors, and adopting a human-centric design approach.

Simulation Assessment of Data-Driven Channel Allocation and Contact Routing in Customer Support Systems

Data-driven operations management methods can transform company operations, respond rapidly to customer demands, and enable new business models. However, companies face the challenge of measuring and evaluating how new technology will impact operational processes. This paper takes a systems engineering approach to assess the tradeoffs of adopting data-driven mechanisms to improve operational processes in a multichannel customer support system. We investigate potential cost savings from two technology applications: classification methods to direct customers to efficient self-service communication channels and routing methods to match customers with agents based on the query type and available skill set. Discrete event simulation evaluates how new technology adoption affects system-level performance. What-if scenarios combine distinct configurations of customer classification mechanisms and available communication channels to evaluate the reduction in the total number of agents required to meet a target service quality level. Discussion includes practical examples of how operational managers could use experimental information to make strategic operational decisions when adopting data-driven technologies.

ОСОБЛИВОСТІ ВПЛИВУ ШТУЧНОГО ІНТЕЛЕКТУ НА БАНКІВСЬКИЙ СЕГМЕНТ ФІНАНСОВОГО РИНКУ УКРАЇНИ

Over the past couple of years, the global economy and financial sector have been increasingly affected by the development of digital technologies. Access to technology, advanced analytics, data and an innovative culture have given FinTech companies advantages that they have used to transform the banking ecosystem. It is these elements that have made it possible to influence access to financial services, payments, infrastructure, distribution and components of the safety and soundness of the banking sector. In today's conditions, artificial intelligence is one of the most important elements that affect the process of digitalization of the economy. With the help of artificial intelligence technologies, the subjects of the banking services market are able to reduce costs and gain remote access to consumers. Innovations related to the use of artificial intelligence enable the transformation of traditional automated remote banking customer service systems into a new technological and informational dimension, resulting in the formation of digital banking. The article examines the impact of artificial intelligence (AI) on the banking sector of Ukraine. The scope of the use of AI technologies in the banking sector, the types and methods of their application are considered. The article analyzes the use of AI in various aspects of the banking sector, including process automation, digitalization of banking services and improved customer service. Among the well-known areas of application of artificial intelligence in the banking sector are chatbots for customer support, risk assessment and credit scoring systems. However, it is emphasized that personalized services have also gained popularity in recent times, as building quality relationships with customers requires not only meeting their current requests, but also anticipating their needs at every stage of the customer journey. As practice shows, personalized sales have a higher conversion rate than selling an accessible assortment of goods for everyone, and this is where artificial intelligence can be applied. The trends in the development of the AI industry in Ukraine and the positive factors of the introduction of new financial services and products on the financial market of Ukraine are also indicated.

E-Commerce Success Factors: PLS-SEM Approach

The current research gap on e-commerce is still relatively incomplete, with predictors in the research model still relatively weak in influence. The purpose of this research is to empirically test the role of the 'satisfaction' variable, whether it is able to fully mediate the relationship between predictors and e-commerce success variables. This research chose the location in the province of West Kalimantan, on the grounds that this province is one of three provinces in Indonesia which has a direct land border with a neighboring country (Malaysia). This research uses the WarpPLS 7.0 analysis tool, arguing that this research is more focused on exploring the 'satisfaction' variable which functions as a mediating variable to 'try' to cover the gaps of previous research. The results of this study found that the satisfaction variable has a significant influence in mediating the relationship between Website service quality, Customer Support System, Electronic Word of mouth and Personalization on E-Commerce success in MSMEs in West Kalimantan province. Consumer satisfaction in being an important factor for business managers, consumers who have high satisfaction will have a positive impact on MSMEs, so that the potential to increase sales turnover becomes bigger and better.

A multiapproach generalized framework for automated solution suggestion of support tickets

Nowadays, customer support systems are one of the key factors in maintaining any big company's reputation and success. These systems are capable of handling a large number of tickets systemically and provides a mechanism to track/logs the communication between customer and support agents. Companies invest huge amounts of money in training support agents and deploying customer care services for their products and services. Support agents are responsible for handling different customer queries and implementing required actions to solve a particular issue or problem raised by the service/product user. In a bigger picture, customer support systems could receive a large amount of ticket raised depending upon the number of users and services being offered. Customer care service gets directly affected due to the high volume of tickets and a limited number of support agents. Therefore, providing support agents with the recommendations about the possible resolution actions for a new ticket would be helpful and can save a lot of time. This study is focused on the development of an end-to-end framework for suggesting resolution actions rather than recommending free form resolution text against a newly raised ticket. To develop such a system, the pipeline is broadly divided into four components that are data preprocessing, actions extractor, resolution predictor, and evaluation. In actions extractor module, we have proposed a technique to identify and extract actionable phrases from resolution text. For resolution predictor, we have proposed two different pipelines that are referred as “Similarity Search Model” and “End-to-End Model.” The similarity search method is based on a ticket similarity search to find the most relevant historical tickets which then leads to corresponding resolution actions. On the other hand, end-to-end model make use of actions extractor module directly and implemented in a way to directly predict resolution actions. To compare and evaluate the mentioned methods on the same ground, we also proposed an actions evaluation criterion which uses BertScore and METEOR score jointly to compute the score against actual and predicted actions for a particular test ticket. The analysis and experiments are performed on the real-world IBM ticket data set. Overall, we observed that end-to-end model outperformed similarity search-based methods and achieved better performance and scores comparatively. The trained models and code are available at https://bit.ly/2GbUBVk.

The synergetic effect of after sales service, customer satisfaction, loyalty and repurchase intention on word of mouth

PurposeUnder a dynamic business environment that is seen today, organizations need to adopt the right blend of strategies to attract, retain and maintain customers. Extant marketing literature has shed light on some key roles that after-sale service and the larger customer support system could play in attaining the same. Specifically, this paper aims to attempt to model the linkages between after-sales service, service quality, customer satisfaction, customer loyalty and word of mouth (WOM).Design/methodology/approachA structured questionnaire was used to collect data from 280 two-wheeler owners, who had a recent experience of after sales service at their respective authorized service centers, through purposive sampling method.FindingsThe findings reveal that both customer loyalty and repurchase intention significantly and positively mediate the relationship between customer satisfaction and word of mouth. This indicates that satisfied customers are more likely to be loyal to the service firms and would spontaneously engage themselves with positive word of mouth.Originality/valueDespite the increasing importance of after-sales service, empirical researches on this subject have been quite limited. This study aims at filling the void in extant literature by examining the inter-relationships between after-sales service, service quality, customer satisfaction and WOM.

A Note on the Level of Customer Support by State Governments: A Mystery-Shopping Approach

Many government services are provided at the state level such as unemployment insurance, Medicaid, and SNAP. Given the lack of competition, a natural worry is that customer support provided by states for these services is less than adequate. While there are many different measures of how a state can support beneficiaries, we focus on just one in this short and applied report: the ability to get a live representative on the phone to help with an application question. To do this, we take a “mystery shopping” approach and make 2,000 phone calls to state government offices. We find substantial heterogeneity in the availability of live phone representatives across states and types of service (UI, Medicaid, etc.). For example, live representatives in New Jersey and Georgia were reached less than 20% of the time while representatives in New Hampshire and Wisconsin were reached more than 80% of the time. We hope that this report provides a simple example for how academics, investigative reporters, and watch groups can help states be more accountable for their customer support systems. Institutional subscribers to the NBER working paper series, and residents of developing countries may download this paper without additional charge at www.nber.org.

A Note on the Level of Customer Support by State Governments: A Mystery-Shopping Approach

Many government services are provided at the state level such as unemployment insurance, Medicaid, and SNAP. Given the lack of competition, a natural worry is that customer support provided by states for these services is less than adequate. While there are many different measures of how a state can support beneficiaries, we focus on just one in this short and applied report: the ability to get a live representative on the phone to help with an application question. To do this, we take a “mystery shopping” approach and make 2,000 phone calls to state government offices. We find substantial heterogeneity in the availability of live phone representatives across states and types of service (UI, Medicaid, etc.). For example, live representatives in New Jersey and Georgia were reached less than 20% of the time while representatives in New Hampshire and Wisconsin were reached more than 80% of the time. We hope that this report provides a simple example for how academics, investigative reporters, and watch groups can help states be more accountable for their customer support systems.

Analysis and safety engineering of fuzzy string matching algorithms

In this paper we explore fuzzy string matching in an automatic ticket classification and processing system. We compare performance of the following string similarity algorithms: Longest Common Subsequence (LCS), Dice coefficient, Cosine Similarity, Levenshtein (edit) distance and Damerau distance. Through optimisation, we accomplished a 15% improvement in the ratio of false positives to true positive classifications over the existing approach used by a customer support system for free customers. To introduce greater safety; we compliment fuzzy string matching algorithms with a second layer Convolutional Neural Network (CNN) binary classifier, achieving an improved keyword classification ratio for two ticket categories by a relative 69% and 78%. Such an approach allows for classification to only be applied where a desired level of safety achieved, such as in instances where automated answers.

Simulation-based Assessment of Data-Driven Processes in Customer Support Systems

The use of digital technology and data-driven techniques in companies has become an engine to transform operations, respond rapidly to customer demands, and enable new business models. However, companies face the challenge of measuring and evaluating how the introduction of new technology will impact operational processes. This paper takes a systems approach to study how to assess the trade-offs of adopting data-driven mechanisms to improve operational processes in a multichannel customer support system. We investigate potential cost savings from two technology applications: classification methods to direct customers to efficient self-service communication channels and routing methods to match customers with agents based on the query type and available skill set. Discrete event simulation experiments model and evaluate how lower-level technology adoption, characterized purely on a functional level, affects system-level performance. Resulting discussions include practical examples of how operational managers could use experimental information to make strategic operational decisions when adopting data-driven technologies. What-if scenarios combine distinct configurations of the customer classification mechanisms and the available communication channels, to evaluate the reduction in the total number of agents needed to meet a target service quality level.