Truth with trust and reputation

Lately, I have been developing a model for some project at my institute. The idea was the crowdsource water point quality assessment in some country in Africa. So users would vote on water quality using SMS. Very basic technology but could make the life of the people that much better by directing them to drinkable water points across vast plains where water and resources are scarce.

 

I have checked the blog post on the meeting results of the swift river team. The whole approach is indeed very good. I think an addition to this approach would be the integration of trust and reputation models into the architecture of the crowdsourced filter. I will explain briefly some working of the model without going into much details here.

 

The model I’m working on is time sensitive which makes it distinct from many other trust and reputation models out there (I will explain what this means later). Also, the model rewards good contributors by a slow build up of user reputations and a fast decline of reputations for those who commit mistakes often. This allows the model to be more resilient to attacks compared to other standard models. 

 

The main idea is that as users make information contributions about some event which are proved to be valuable to the community, the reputation of the user will grow. This is some form of a rewarding system. Research shows that a major factor of encouraging users to contribute in crowdsourcing is to acknowledge the people’s contributions and provide some sort of reward that does not have to be material, but in the majority of cases the rewards are in the form of gratitude. The reputation system rewards users while giving them the opportunity to build AND maintain their reputations in their community. Their reputations are then used to filter the information contributions they make in the future, making it possible to develop a crowd sourced filter that does not on the longer run require much intervention from information assessors as opposed to information contributors. This is one aspect I think is lacking in the current Swift River approach. The system is always dependant on having a crowd responsible for assessing information, while the system I propose is self learning and self sustaining which makes for a good compliment to the current Swift River filter architecture.

 

The diagram below shows the performance of the model for rewarding good contributions. It basically shows how the reputation of the users changes after good contributions to the system. The higher the reputation of some users the higher the veracity of future contributions from these users. It is safe to make that claim, since trusted media sources (blogs, tweets, etc of people) are simply assumed to consistently provide trusted information.

 

 

           

 

On the other hand fraudulent or lousy folks will be subject to the behavior in the diagram below:

 

 

 

 

I

I

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Information triage using a reputation and trust based system is proven to be robust, and if the model is resilient to attacks then the information filtered using the model is certainly more trusted.

 

another interesting feature of the model is that the reports made by the users (say a tweet for e.g.) are given trust ratings. Such trust ratings are subject to time decay themselves. So that a report that is 5 hours old loses trustworthiness as time passes in favor of newer more trust worthy reports. Such behavior, of course, can be turned on or off. But it comes in handy in cases of high volume high intensity information situations where the information value decays over time. This is the case with the water point assessment because in this African country, a water point that is good today might no be good tomorrow. The diagram below shows how the model deals with this problem to provide final status values for water points over time (or any other information entity for that matter).

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

The model, of course, can and will be extended to include spatial aspects of the information, such that proximity of the source to the events can be taken into account as well. I have also developed a separate model which just deals with the spatial aspect, which is a simple model, but the foundation of it remain unchanged, and the model can certainly be further improved.

 

I think integration of a time sensitive trust and reputation model in the Swift River architecture can lead to a more self sustaining system and more automated filtering as the system learns more about its users.

 

any comments, ideas? 

 

 

ChaCha and Trust

A while ago I stumbled upon ChaCha which has a unique claim to fame. ChaCha allows you to perform standard search, but also to perform human assisted search! You simply create an account and ask for guided search, and a human on the other side (an actual human that is) assists you in finding the best search results. Another service provided by ChaCha is mobile search, where you send a question and the human guide sends you back an SMS with the requested search results, so you can find a specific restaurant  with a specific rating in any place with really minimal effort.
The idea did occur to me before to be honest, but I ridiculed myself for being outlandish! Imagine people doing the search for you on the other side! with well trained cheap labour in a remote country it seems like ChaCha did it, I’m not sure though it is cheap labour and I certainly hope they are being fairly paid.One would of course wonder how can you trust a remote person to provide you with results, I personally find it hard to accept and would mostly prefer doing my own search, specially that I’m good at it.
Would you trust a complete stranger to do a search for you? ChaCha’s main challenge in my view is to foster a community of trust. ChaCha users and the search guides need to develop a unique form of trust that enables them to work together for the good of the users. At the moment I find it hard to figure out what are the elements of a trust model that would be valid for a service like ChaCha. Can we make this trust model explicit in how ChaCha functions? How will users develop trust in complete strangers to perform searches for them and select whatever results that are relevant for a ChaCha user?.
To my mind if ChaCha is to be a success and I think it will be, I think understanding the basics of trust and how users assess their interactions with the search guides is key to the future of ChaCha. 

Written From: office, Weather: so damn cold, Mood: energetic but worried

Geospatial Social Web? Lets invent stuff!

Sometime ago a term surfaced in the geospatial domain. It was “Geospatial Semantic Web”, whatever that means! One can assume it has to do with the application of semantic web technology to geospatial web applications, but you can never really be neither sure nor specific.

At the start of my research work I was very interested in the new social technologies collectively and widely termed Web 2. I thought there was hugh potential for the geospatial domain. I published a paper in AGILE 2007, which was just exploring possibilities of work and the paper was well received. I didn’t dare however to invent yet another new term such as that hanging up there as the title of my blog post. I thought one needs to be more conservative about inventing terms haphazardly if you really wanted to do serious science.

During the EC-GIS in Porto, an event which I didn’t attend but our colleague Eva Klien did attend. Someone in that meeting seemed to have used the term SDI 2.0!!! ain’t this funny? Scientists are acting much like freelance bloggers and internet geeks, just inventing terms without specific well defined meanings. Not a rare event in science (some sciences, not all) which I was made to view with a more careful eye in the face of such a nice article which I came to read lately. This article is a highly recommended read for many in the computer science/informatics fields. I do agree with the point the writer is trying to make and this point is very apparent in the Geoinformatics field, it is namely that there is lots of pseudoscience and not-so-scientific-work going on and we need to improve a bit more on that.

So why do I use the term “Geospatial Social Web” on my post title? well, my logic is that if someone dare invent such a term, I’d have the credit of being the first one to have mocked it, pre-invention! You do see the logical fallacy don’t you?

Written From: office, Weather: sunny intervals and warm, Mood: energetic and calm

Oracle 11g, Spatial support expanded and OWL ontology support too!

over at the location intelligence blog there is a very interesting post about Oracle 11g improvements in spatial support. Interesting enough in web services infrastructure they support WFS 1.1 nd CSW 2.0 in addition to the already supported WMS 1.1. This is great news for OGC standard based service implementations.

One other note over at the ebiquity blog has to do with OWL support in Oracle 11g. It will be possible to do inferencing for a subset of OWL on Oracle. this means you can build queries enhanced by some OWL axioms or queries. So much for OWL into the commercial world. I expect those will be primarily targeting biotechnology markets where OWL based systems seem to be gaining more momentum.

Googlemaps API with KML and GeoRSS support

Interesting news, been around for a few days though! Google have enabled KML and GeoRSS integration into Google Maps API. This cool stuff is announced today on the googlemaps blog since it was badly needed, Also geoRSS was already supported by Yahoo API for a long while now!

GeoTagging Foot Prints.

Geo-tagging is already part of mainstream tagging. I have been thinking of implementing something that resembles regions that are derived from geotags. Simply one can use a foot print algorithm to create regions out of points. The points in this case are the geotags. And the resulting regions can be returned when the geotags are queried. The benefit is that those regions will have all the relevant tags enclosed and one can difrectly see on a map how people are tagging a certain area. You are easily able with such mapped tag foot prints/regions to figure out for example that a certain region of down town Paris has so many restaurants or pubs. This would be useful don’t you think?. Moreover people can start to tag the resulting regions and make them even more well defined.

Written From: home, Weather: rainy and windy, Mood: energetic

Mapping Social Networks

It is interesting how web social networks let you make new friends with minimal effort. I wonder if these are friends in the conventional sense though. They are more like e-pals. Research on social networks shows that geographic proximity is probably the best determiner of how often friends get together. It would be interesting of online social networks are sensitive to the location of it’s members to provide more geographic based functionality and categorization. It makes sense then that those who are geographically closer will tend to trust each others more over time as an artifact of more interaction. If social networks are location aware maybe they can eventually help us make more real friends out of our e-pals. The question remains, which is better? friends or e-pals? e-pals or friends?

Written From: office, Weather: sunny and temperate, Mood: jumpy