Shopster – A Social Mobile App

Recommendation system is one of the most important features of Shopster.  Recommendations will be proposed to a user when he is viewing an offer and when he is searching for . There are many types of recommendations that our system will recommend. Some of them are (lets assume that the reader is viewing an offer about a laptop):

1. Offers about related products (e.g  the system will recommend offers about laptop cases, a mouse, etc. )
2. Offers about similar products (e.g the system will recommend other offers about laptops)
3. Other offers from the current shop
4. Offers liked by other friends (offers that friends of the user have rated highly)

A more detailed analysis of the recommendation system and algorithms to be used can be found here.

Every user will be able to rate the offers presented in our application. The combined rating from all offers of a user will be his/her rating, so other users will know if a he/she is a good/reliable one in order to trust him/her. The ratings will be out of 5. We can still make some simple categories to rate, like value for money or purchase happiness and then calculate the overall rating. This rating will show if an offer can be trusted or if is a good deal for other users. Furthermore with this rating system we will be able to rate each user who posts offers, making clear to other people who is a good/trusted user.

In order to make users contribute more to the application, rewards will be offered to them. Rewards could be in form of a coupon that gives a discount in a specific shop. Coupons will be given to users based on their contribution, for example how many friends they invited, on the number of their posts and comments and on their rating.

As we can see in this post (http://www.readwriteweb.com/archives/how_to_make_your_location-based_app_a_success_reward_people_for_activities.php), rewarding users in a location-based mobile applications is the new key to success.

Users could be categorised according to their number of posts. For example a user without any posts or a limited number of posts is considered novice. Users with many posts are categorised as expert users. More categories of users could exist between expert and novice. Also users can rate the comments or posts of other users, and the ranking of a user could be calculated based on posts rating, comments rating, number of posts and number of comments.

A user will be able to make an account and register to Shopster. The Then he/she will be able to login and use the application. The user will have to provide some basic personal information in order to create his/her account. These will be: username, password and email.

Except from that, a user will be able to connect to the application using his existing  Facebook or Twitter username and password. This is a good thing for many people who do not want to give their personal information to every single site the sign up.  We have already found some simple ways to implement this:

• For Facebook integration here : http://www.android10.org/index.php/articleslibraries/290-facebook-integration-in-your-android-application
• For Twitter integration here : http://www.android10.org/index.php/articleslibraries/291-twitter-integration-in-your-android-application

The users of the application are divided into two categories: simple users and shop owners. Simple users are ordinary application users who use the application for viewing and sharing offers and discounts from third party sources.

In contrast to simple users, shop owner user category is first-hand information source whose members can use the application to input the offers and discounts available at their stores thus generating new content and can share it with other users. Thus the application provides a platform which the shop owners can leverage to profit their business.

The application recognises the different categories of users, and provides features accordingly.

One of the essential features of the Shopster application is to let the users post new offers and discounts and share it with other users.

For example, if an ordinary user comes across a new discount or offers in a particular shop, and decides to share it with other users (or a specific group of users), he can easily do that using this application. The application provides the GUI to input the relevant information (location, expiry date of offer, visibility of the post etc) and with one click the information will be available for other users to view!

A different scenario is when a shop owner decides to post the offers and discounts related to his shop with profit motive. He can use the application similarly and post the information and share it with other users (Here the post will be mostly public).

Registered users of Shopster will be able to submit their comments on the offers posted (apart from rating them). This way they will have the chance to express their opinions and say whether an offer is worthy for example, or maybe say if there is a catch for obtaining a specific offer, or even give some details about the technical specificiations of a product. However, these comments should also have some kind of evaluation and undergo criticism from the rest of the users reading them. Therefore, we have decided to include a rating for every comment and other users will be able to say if that comment is helpful or not. These ratings will then be weighted and will reflect to the user’s overall rating. So if for example a user’s overall rating is quite low and a new Shopster user sees a comment from that user, he/she will know that the specific user is not very trustworthy and that the comment is probably not accurate.