Journal Articles

Permanent URI for this community

https://ir.gretsauniversity.ac.ke/handle/123456789/253

Browse

Browsing Journal Articles by Subject "TECHNOLOGY::Information technology"

Now showing 1 - 1 of 1
  • Thumbnail Image
    Item
    An Ngram-Based Approach to Determine Trends and Patterns in the Social Networks
    (IEEE AFRICON, 2023) Constance Mukina Ngila; Waweru Mwangi; Michael Kimwele
    The recent progress in computing has made it easier to collect and store huge amounts of information in a text. The growing size of text datasets in text mining and the high dimensionality associated with knowledge discovery is a great challenge that makes it difficult to classify documents into various categories and sub- categories. This paper focuses on how text can be mined from social networks and then categorized using n-grams to determine specific trends and patterns. The main aim of Knowledge Discovery is to extract knowledge from data in the context of large databases. The volume of information that is available is increasing every day. This data ranges from that used in business transactions to scientific data, sensor data, pictures, videos, etc. There is, therefore, a need for a system capable of extracting the core of available information and automatically generating reports, opinions, or summaries of data to aid organizations in better decision-making. Knowledge Discovery is a repetitive process where evaluation measures are often enhanced, mining done on data can be refined, there is an integration of new data, and the data is transformed to get accurate and more appropriate results. The data collected from social networks need to be filtered to capture specific text that will be useful to a PR brand following what clients say about their products online. There is a need for a technique that will provide a quick and precise way of fetching specific text from huge amounts of data on social networks to help analyze the feedback. This research analyzes the use of ngrams to fetch specific text from near-real-time customer feedback that is in the form of large data on Twitter to help Public Relations agencies determine the trends and patterns that will help them align their brands with customer preferences. Keywords—knowledge discovery, data mining, trends, and patterns.