Framework for Predicting Customer Sentiment Aware Queries and Results in Search Using Oracle and Machine Learning

Abstract

Users today need to express their informational need in a way such that the search results can be further analyzed to directly address the need, instead of merely returning a list of lexical hits. The rise in popularity of machine learning (ML), and deep learning in particular, has both led to optimism. In the industry analyzing customer feedback is very important to improve service quality, identify and troubleshoot key networking areas, enhance user experience and provide them with better quality. Traditionally, sentiment analysis is performed using external machine learning frameworks, leading to integration challenges and performance inefficiencies, first huge amount of data is transferred through different options like API calling or FTP file sharing then these data machine learning models are applied to get data insights. In this paper, we address the problem of extracting sentiment metadata related to the user’s topic or entity of interest along with the search results. A real-time sentiment analysis model implemented directly within Oracle Autonomous Database features Oracle Machine Learning (OML4SQL and OML4PY) to classify telecom customer feedback collected from a web-based Oracle APEX feedback system. The study utilizes Oracle’s built-in machine learning algorithms, including Naive Bayes and Support Vector Machines (SVM), to train a predictive model for classifying customer reviews as positive, negative, or neutral. The trained model is then applied to new feedback, enabling real-time sentiment prediction without external AI platforms. This avoids API’s or FTP file sharing and all the processes completed within Autonomous Oracle Database. The results demonstrate that OML4SQL and OML4PY can effectively classify customer sentiment, allowing telecom companies to gain actionable insights, improve customer support, and enhance decision-making. This research highlights the potential of Oracle machine learning (ML4SQL and OML4PY) for real-time text analytics in the telecom database, Opening the door to automated, sentiment-based decision-making in customer service.

Keywords: Sentiment Analysis, Oracle Machine Learning for SQL (OML4SQL), Oracle Machine Learning for python (OML4PY), Telecom Industry, Customer Feedback, Natural Language Processing, Naive Bayes, Support Vector Machines, Oracle APEX Web-Based Sentiment Analysis.