The differences between a scenario and a use case

Posted on April 28, 2015 by Jingyuan Li

Use cases are a technique for capturing the functional requirements of system.
Use cases work by describing the typical interactions between the users of
a system and the system itself, providing a narrative of how a system is used.

Rather than describe use cases head-on, I find it easier to sneak up on them
from behind and start by describing scenarios. A scenario is a sequence
of steps describing an interaction between a system and a user. So if
we have a Web-based on-line store, we might have a Buy and Product scenario that
would say like this:

The customer browses the catalog and adds desired items to the shopping basket.
When the customer wishes to pay, the customer describes the shipping and credit
card information and confirms the sale. The system checks the authorization on
the credit card and confirms the sale both immediately and with a follow-up email.

This scenario is one thing that can happen. However, the credit card authorization
might fail, and this would be a separate scenario. In another case, you may have
a regular customer for whom you don’t need capture the shipping and credit card
information, and this is a third scenario.

All these scenario are different yet similar. The essence of their similarity is that
in all these three scenarios. the user has the same goal: to buy a product. The user
doesn’t always succeed, but the goal remains. This user goal is the key to user cases:
A use case is a set of scenarios tied together by a common user goal (Fowler, 2004).

scenario diagram

http://www.inf.unibz.it/~ryzhikov/HCI/hci_02.pdf

References:

Fowler, M 2004, UML Distilled : A Brief Guide To The Standard Object Modeling Language / Martin Fowler ; [Forewords By Cris Kobryn … [Et Al.]], n.p.: Boston : Addison-Wesley, c2004., University of Southampton Library Catalogue, EBSCOhost, viewed 28 April 2015.

Game Theory and recommendation systems

Posted on April 25, 2015 by Muna Al Abri

Recommendation systems are quite widely used in the web. Amazon use it to recommend books, youtube to recommend videos based on other users preferences who are similar to the active user. Recommender systems helps users to express their interests. This has helped massively on the explosion of information on the internet. Imagine if Amazon didn’t have a recommender system, users won’t be able to be exposed to the same amount of goods they are currently exposed at.

Small.World will be using a recommender system that will help matching friends who share other friends and interests. (Azadjalal, Moradi and Abdollahpouri, 2014) came to a conclusion that most people rely on recommendations by friends rather than by strangers and this is our emphasis on Small.World. Recommendation will be made on how much the active user trust others.

The main concern is how to improve this matching ? To answer this question, let’s take a look at one of the most important features of Recommender systems which is “Collaborative Filtering” that predict ratings for active users on things/people based on other user’s opinion. Those ratings can be improved using trust statements. The paper suggests a novel method to improve trust in Recommendation systems using game theory (Azadjalal, Moradi and Abdollahpouri, 2014). Using Pareto Dominance concept which is a state of allocation of resources where making an individual in the network better is impossible without making everyone else on the same network worse-off (Wikipedia, 2015). the concept doesn’t take equality under consideration. It is applied to income distribution where the rich gets richer and the poor gets poorer.

Game theory is used to help identifying those trustable users who correctly have the same interests of the active user and who therefore should be the candidate [FOAF]. The method suggested applies pre-filtering process that reduces the chances of choosing the least trusted users from those who are trusted. To do that, the method uses a process with five steps to predict trust for each user as the fig.1 shows.

Figure 1. Algorithm of the proposed method (Azadjalal, Moradi and Abdollahpouri, 2014)

Trust is a hot topic on the social networks. Algorithms for trust in recommendation systems are developed to help users make wise decision. Small.World tends to use the proposed algorithm because of the matching requirements. Also, Small.World tends to adopt the latest and most efficient algorithm if it will meet our needs.

Refrences:

Azadjalal, M., Moradi, P. and Abdollahpouri, A. (2014). Application of game theory techniques for improving trust based recommender systems in social networks. In: The Fourth International Confrence on Computer and Knowledge Engineering. [online] Sanandaj, Iran: IEEE Computer Society, pp.261-266. Available at: http://ieeexplore.ieee.org [Accessed 24 Apr. 2015].

Wikipedia, (2015). Pareto efficiency. [online] Available at: http://en.wikipedia.org/wiki/Pareto_efficiency [Accessed 28 Apr. 2015].

Are you trustworthy ?

Posted on April 13, 2015 by Muna Al Abri

Trust is involved in human activities in a daily basis. Measuring trust in a digital world can be more complicated than real life. There are various methods that help in calculating trust and evaluating users based on trust value.There are significant number of models in trust in social networks. Small.World tends to follow reputation-based trust. This comprises two type of trust:

Global Trust: it is a quantitative approach where each user scored a value. The global trustworthiness then based on these scores.
Local Trust: it is a qualitative approach where the trust based on personal bias (Ziegler & Lausen, 2004).

The Local Trust are following uses three properties of trust in social networks:

Transitivity : If A trusts B and B trusts J then A will also trust J

Asymmetry : If A trusts B it does not follow that B trusts A

Personalization :Trust is held from the view of a node – it is therefore local trust (Millard & Imran, 2015).

the usage of Online Social Network (OSN) trust methodologies can boost the privacy and security for the user. A user needs to know if they can trust someone to look at there profiles or how much they can trust them. The post discuss two methods: Trust-Online Social Network (T-OSN) and Trust Indexing algorithm (TI).

The evolution of OSN encouraged people to meet strangers from all over the world and become friends. But this has also, led to some problems such as identity theft and stalking. Methods that calculate trust and help users to make decisions are usually complex or require huge resources. T-OSN is a trust model that has been created specially for OSN that can also be used on other applications such as mobile message forwarding and peer-to peer file sharing network (Li and Bont, 2011).

the model is effective, simple and require the least resources possible. the model is based on two things :

1- Number of friends (Degree)

2- Contact frequency (Contact interval)

and then a user trustworthiness can be calculated for each user. The theory suggests that if a user has more friends and more frequent communication with friends, then this user is highly secure to interact with. So the user will get a higher trustworthiness.

Another method is called “Trust Index” (Tang et al., 2012). It has the same idea as Page Rank that is being used in Google search engine. The algorithm is based on distance that counts hops. How far the user from his/her peer.

Each user will have a TI and a list of the ranking of other users, those who are more trustworthiness will be on the top of the ranking. The paper claims that TI successfully been able to classify trustworthiness users from those who are not (Tang et al., 2012).

Google have introduced Google+ circles where recommendations of people where based on common interests or people that the user might be connected to (Tang et al., 2012). Spam was not an issue. Small.World will follow the same path where trust will come in top of the recommendations that the users will get.

choosing a method to evaluate can be tricky where there is no ultimate solution. It depends on the type of the OSN and what factors will be involved in the algorithm.

research on this area is highly encouraged to create a more secure environment for OSN.

References:

Millard, D. and Imran, M. (2015). Trust: Part I.

Li, M. and Bont, A. (2011). T-OSN : a trust evaluation model in online social networks. In: IFIP Ninth International Conference on Embedded and Ubiquitous Computing. [online] Melbourne, Australia: IEEE Computer Society, pp.469-472. Available at: http://ieeexplore.ieee.org/xpl [Accessed 12 Apr. 2015].

Tang, R., Lu, L., Zhuang, Y. and Fong, S. (2012). Not Every Friend on a Social Network Can be Trusted: An Online Trust Indexing Algorithm. In: IEEE/WIC/ACM International Conferences on Web Intelligence and Intelligent Agent Technology. [online] Macau SAR, China: IEEE Computer Society, pp.280-285. Available at: http://ieeexplore.ieee.org [Accessed 12 Apr. 2015].

Ziegler, C. and Lausen, G. (2004). Spreading activation models for trust propagation. IEEE International Conference on e-Technology, e-Commerce and e-Service, 2004. EEE ’04. 2004.

This post is written by : Muna & Awezan