Page 201 - The-5th-MCAIT2021-eProceeding

Preliminary Work on Bag-of-Requirement Representation for
SMA Reviews

d
b
a*
c
Mustafa Abdulkareem , Sabrina Tiun , Umi Asma` Mokhtar , Masnizah Mohd
a,b Asian Language Processing (ASLAN) Lab, CAIT, Faculty of Information Science and Technology, UKM, Selangor, Malaysia
c Information Governance (IG) Lab, CYBER,Faculty of Information Science and Technology, UKM, Selangor, Malaysia
d Cyber Intelligence (CI) Lab, CYBER, Faculty of Information Science and Technology, UKM, Selangor, Malaysia
*Email: P93823@siswa.ukm.edu.my

Abstract
Information extraction from user reviews has recently been used for social requirement retrieval to enhance the
social media apps (SMA) development process. However, with the overwhelming amount of reviews, one needs
to identify reviews that are relevant and ignore reviews that are not relevant or non-informative. To facilitate in
identifying the relevant reviews, a representation related to social requirement terms of such review seems
necessary. This paper describes our on-going work on building the SMA reviews representation by combining
the Wordnet lexical and review word embedding and relates the representation to the standard social
requirements. The work is still at the very early stage where we layout the method to combine the word
embedding and Wordnet, expanding the social requirement term, and relating review to social requirements.

Keywords: Social requirement, Social media application, social media reviews, Word embedding, WordNet;

1. Introduction

Nowadays, social media applications (SMA) such as Facebook, Instagram, Twitter, etc., play a large part in
people's lives, such as sharing feelings and their daily activities Continuously, which makes this information
easy to access and in a simple way for anyone (Hemmatirad 2020). The sharing of information by the massive
number of users has led to the emergence of many text classification techniques, opinion mining, and
information extraction (Martinez-Rodriguez, Hogan, and Lopez-Arevalo 2020).
Popular application platforms such as Apple and Android allow users to evaluate applications by writing
texts that refer to personal reviews about the application. These reviews are among the rich sources for software
developers to upgrade their applications to users' needs. Also, these reviews are of high importance for new
users.

2. Research Gap and Related Work

Existing research on mining app store reviews has been focused on extracting and classifying technically
informative reviews into bug reports and feature requests (Alomar et al. 2021). However, little attention has
been paid to extracting and synthesizing the social requirements present in user reviews. Social requirements,
such as security, privacy, and performance, enforce various design constraints throughout the software
development process (Whitworth 2011). Addressing these constraints is a critical factor for achieving user
satisfaction and maintaining market practicality.

E- Proceedings of The 5th International Multi-Conference on Artificial Intelligence Technology (MCAIT 2021) [188]
Artificial Intelligence in the 4th Industrial Revolution