Amancio Bouza's Research Blog » Social network

Applicability of Social Network Graph Patterns to Recommender Systems

Amancio Bouza — Sun, 24 Jan 2010 09:38:58 +0000

User interface of the spotting application

Some research has already been done investigating Web-based social networks and its applicability for different tasks such as trust inferrencing with trust networks or collaborative filtering respectively recommender systems. My master student Reto Hodel applied social network analysis to a social network and used the metrics to predict item ratings.
For this purpose, he built a very nice looking Web-based location recommender system called spotting. In this application, people get many locations presented on a geographical map that is based on Google Maps. The meaning of the encoded locations is very intuitive. The size of the location determines the level of match between the person’s preferences and the location. The bigger the location is presented the more relevant it is.
In order to boost the process of social network developing, we decided to build uppon an existing social network. Thus, we can rely on a already existent social network that has developed the relations among people already. For this reason, we decided to use the Web-based social network of Facebook because it provides a useful API, the Facebook API. Instead of developing a Facebook app, we decided to develop a stand-alone application that uses the Facebook Connect provided by Facebook. Facebook Connect is a function that a person can use to login to a different Web page with his Facebook account information. With this function it is still possible to gather his public information from Facebook, such as personal information and friends.
The value respectively usefulness of a collaborative recommender system increases with the amount of people and ratings. Therefore, the theory of network externalities respectively network effects applies to such systems. Thus, attracting people to use and rate the application in the first place is crucial.
Besides the social network and API’s that are provided by Facebook and others, Facebook also provides the facility of publishing news on the person’s wall that is seen by his friends. We used this function to publish location ratings and attract his friends to use spotting too. Our experience showed, that even with relying on a existing and developed social network it needs a lot of marketing to attract people. People do not just use an application because their friends use it. In general, no general approach exists to overcome this problem. First attempts try to reduce the so-called cold-start problem that describes the situation where only few ratings exist. But this is just one part of the solution.

Social Network of spotting

Anyhow, we could get 139 people in 3 weeks. Their ratings were the basis of our analysis. We analyzed the metrics of trendsetter, cliques, friend chains and some others. Unfortunately, we could not show that one social network metric leads to more accurate rating predictions. In contrast, the location’s average rating has been shown to perform better or at least equal to the social network graph patterns that we have investigated. In addition, the average rating is very cheap to compute! We applied the Friedman test that gave us evidence that some significantly performance difference exist on the significance level of 5%. Therefore, we run the Wilcoxon rank test. On the given significance level, we could not determine one single approach that is better then all the other. Of cause, we applied the Bonferroni correction for the family-wise error. The predictions based on trendsetters have been shown to be worst.
But these results have to be taken with caution. Despite applying statistical tests, the data set it self has some major threats to validity. The social networks consists mainly of the personal social network of the master student. That means that the average rating consisted mainly of his friends’ ratings. This fact highly influenced the performance of all other social network patterns.
However, I think that the experiment shows that the simple concept of wisdom of crowds respectively average rating is a simple but effective approach to provide ratings. Further, the average rating has the highest item coverage since patterns such as cliques, trendsetters, etc. do not give that much information about that many items.

Error distribution of cliques

To conclude, social network patterns may provide valuable information to generate more accurate recommendations, but not in general. The high computational costs of computing social network metrics should be taken into account because some of them are NP-hard.
However, other researchers investigated more positive results. But their evaluations should be read rather carefully. Sometimes, they tweak the experimental setting by defining weak hypotheses and unit of analysis to favour the social networks.
Thus, just be careful when people sell their work on social networks and be aware of the structure of the social network itself, that influences highly the experimental results. For instance, not all Facebook friends are real friends in Real-life.

Abstract

Generating accurate recommendations for items, such as locations, movies or books, is challenging. Common Web-based recommender systems require information about the users’ past to generate suitable recommendations for them.
In this thesis we first present spotting.li, a location recommender system based on Facebook, which allows users to rate locations and generates recommendations inferred by their friends’ ratings. In doing so, we examine requirments to successfully implement such a system using the latest web technologies (i.e., Grails) and describe key elements of our approach. Our focus is put on performance and providing an easy-to-use interface incorporating Google Maps.
Furthermore, we analyse different recommendation approaches which leverage structural in- formation from a social network to predict ratings. In particular, we examine the use of social network patterns, such as cliques and trendsetters, as well as direct friends and two levels of indirect friends. We finally conduct an extensive evaluation of these approaches, based on real data collected during the time of the thesis.
To prove our findings, we test our dataset, based on 139 users, for statistical significance. We demonstrate that even a simple algorithm, such as the average rating, bares similar results to more elaborate algorithms.

Reto Hodel: “spotting – Realisation and Analysis of a Location Recommender System Based on Facebook“, ed. by Amancio Bouza and Harald C. Gall, University of Zurich, December 2009. (master thesis)

spotting – Social Network-based Location Recommender System

Amancio Bouza — Thu, 29 Oct 2009 18:07:16 +0000

User interface of the spotting location recommender system

Several weeks ago, my student Reto Hodel who I’m supervising currently published a really cool and easy-to-use Web application that lets you find new potentially interesting locations such as bars, restaurants or clubs based on their social relationships within the Facebook community. In other words, your recommendations are generated depending on your friends, your different participating n-cliques, your explicit and implicit trusted trendsetters and other social network analysis based metrics. This Social Network-based location recommender system is called spotting.

In this master thesis, he built a Web applications that uses the people’s Social Network at Facebook to generate the location recommendations. Based on the data gathered by the Web application we analyze what SNA metrics are suitable for generating such recommendations. We are still investigating this research questions. As soon we get som e more insights we plan to add them to the Web application such that people can benefit from better recommendations.

If you are interest in having a look into the Web application then please just check out the following link:
http://seal.ifi.uzh.ch/spotting

You may want to join the spotting group at Facebook. So far, we only provide locations from Zurich.

My first Bachelor student finished his thesis about “Purple Leaf – Evaluation of the Adoption of New Features in a Web-Based Social Network”

bouza — Sun, 04 Jan 2009 20:14:43 +0000

Participants resp. party people on the various partys

Together with Marc Vontobel’s main advisor Gerald Reif, I successfully advised and supported Marc with his bachelor thesis. He was my first student I advised at all. He reenginered Purple Leaf – a party portal – and investigated the impact of social relations among people with respect to acceptance rate of new features on a web page. At the time he finished with his thesis, the community of Purple Leaf consisted of several thousands of people.
It seems that one rather accepts and adopts a new feature when a friend already adopted it. It seems that the inhibition threshold is much lower due to the fact, that people trust their friends and thus, trust the feature and it added value. Marc did a great job that could not be expected from a Bachelor student, starting from the software engineering skills to the data gathering and complex Social Network Analysis (SNA). But he was already prepared from our precedent seminar on “Trust and Recommendation in Social Networks”.

We were surprised as he delivered not only his thesis but also a complete picture book with fantastic analysis of differenc aspects of his party community. He discovered the core group of party people being on most parties, the latent semantics among different music styles and drinks.

Abstract

Purple Leaf is a social network which offers its member several possibilities to personalize its exclusive events by providing them unique online services. After the size of our platform suddenly increased from 300 initially invited guests to a multiple, we were obliged to completely revise the platform and enlarge our range of services. To embed these new services smoothly into the existing web presence, we fully restructured the application and changed the basis to a modern web framework. After that makeover, we designed five other services which we targeted to increase the customer loyalty and the entertainment value of our platform. Because new features are often not instantly accepted by existing users, we developed an integrated concept for boosting the acceptance of novel functionality. This concept is based on the technology acceptance model which was developed by Davis (1986). The model postulates that the actual use of a new feature is solely based on external factors. On the one hand, there are factors which influence the ‘perceived ease-of-use’ and on the other hand some that have impact on the ‘perceived usefulness’. In order to foster the perceived ease-of-use, we developed several usability concepts and tried to figure out how Web 2.0 features can help to simplify different processes. Beside the creation of intuitive user interfaces and plain procedures, we worked on an elaborated data and application structure which itself also contributed a big part to the simplicity of the new functionality. After we had embedded the services into our Internet portal, we started to analyze the acceptance of one new feature: ‘The most favored Guest’. This service allows every sign up member to define his personal list of favored guests for an upcoming event. Once the selected users are informed about their election, they, in turn, have the chance to define their own list. After a first round of selection, we tried to boost the personal acceptance of our members by providing specific incentives. Beside the active interventions into the process of adoption, we also analyzed a passive phenomenon: Does some kind of peer pressure exist within virtual cliques? If so, there might emerge some interesting changes in common marketing strategies which could narrow down the target audience to some single users of the network. In addition, we visualized some of the encountered situations and putted them together in an illustrated book as supplement to this paper.

Marc Vontobel: “Purple Leaf – Evaluation of the Adoption of New Features in a Web-Based Social Network”, ed. by Gerald Reif, Amancio Bouza, Harald C. Gall, University of Zurich, December 2008. (bachelorsthesis)