Eulogy Writing Using Social Media Activities

Abstract

Funeral speech is an opportunity to present tribute to the decreased person by giving a short speech about their life turnings and successes. It’s an honour for the loved one or close friend to present tribute to the decreased person with dignity and honour. Eulogy or funeral speech is a short piece consisting of few lines and covers the whole life of the dead person. Most of people are busy in their personal lives and have lesser time to spend with friends and family. The physical connectedness has been replaced by electron communication through social media sites such as Facebook, Twitter, and WhatsApp etc. While if we don’t know much about activities of loves ones and friends, one possible choice is to gather information about some particular person using social media websites. Through access of free texts from Facebook posts and Tweets etc., this is possible to have deep details about any person. The social media text is mostly the unstructured text mixed with images, emojis and contextual details. This type of unstructured data require experts’ knowledge to extract useful knowledge through manual annotations. Alternatively, the Natural language processing (NLP) and text mining techniques bring the best from those free texts. Using NLP, machine learning and data mining techniques, it is possible to collocate useful information about a person and develop a perfect eulogy statement. This research performed Tweets text data collection from Twitter API in R through Rtweet package. Pre-processing techniques were employed to create a clean corpus from Tweets. To understand the most important events of a personal life, topic modelling techniques from NLP was applied on the processed corpus. Latent Dirichlet Allocation (LDA) technique was applied on a sample of 10,000 tweets due to computational complexity. Results assured that the use of social media for eulogy statement writing is new and easier approach.