Group recommendation systems are designed to make recommendations to groups of users based on their shared interests and preferences. [Dara et al., 2020] enlists the prominent existing group recommendation techniques.HappyMovie[Quijano-S´anchez et al., 2011] is an example of a system that recommends movies to groups of users on Facebook, using social and behavioral information to enhance recommendations. LGM[Shi et al., 2015] and other systems [Boratto et al., 2010] use automated approaches to detect groups based on shared interests and preferences. Probabilistic approaches ([Gorla et al., 2013], [Yuan et al., 2014]) ,  estimate the probability that users will like a certain item and use group relevance to produce higher quality recommendations. Hybrid approaches([Baltrunas et al., 2010], [Kaˇsˇs´ak et al., 2016]) use collaborative filtering and content-based approaches to generate group recommendations. Some systems, such as GrAM and Hungarian Aggregated Method [Agarwal et al., 2017], consider the order of user preferences to calculate satisfaction.

However, all these works have their own pros and cons. To the best of our knowledge, there does not exist an entire pipeline which takes the pros of the existing techniques to generate efficient recommendation. Our proposal is to make use of the existing techniques to create an optimal pipeline that will provide very efficient group recommendations. Also, there does not exist a system which analyses the effect of one technique on another. We planned to explore this as well.

1. What is the relative performance of various combinations of group recommendation techniques when compared to one another?

2. Is it possible to get an optimal pipeline for group recommendation techniques, based on efficiency or performance?

3. Based on the comparative analysis to find the optimal solution, can we also find the factors/techniques contributing more to optimality and the ones contributing less or deviating from or hindering optimality?