Services
 |
STAMPSSystem for TAgging Messages, Post-Inferential Semantics |
Purpose | Approach | Publications | Contacts | CRAFT | BLOG | Field Trial
Abstract
This work targets specifically a human activity that we defined as Collaborative Annotations of a Map in a mobile setting. In general, this can be seen as a form of `Spatialised' communication, which means that communication makes explicit usage of the geographical/physical context as referent to the message content. The goal of this study is to develop modeling schemas that enable to integrate spatial information, as embedded in maps, with the textual information produced through computer-mediated communication. Therefore the expected outcome of this study is twofold: on one end we plan to find a computational description of space using the annotations of the space inputed by the users (and viceversa); on the other end, we are planning to proof the goodness of such description through an applied verification.
Research question
In this research, we will develop modelling schemas that enable to integrate spatial information, as embedded in maps, with the textual information produced through computer-mediated communication. This integration will be investigated through the development of a specific search engine that combines these two sources of information. In other words, instead of offering a mere word-matching search, it is our intent to develop a more advanced mechanism which will take advantage of the spatial distribution of the messages, the threading between them and the user's interaction in the system. Therefore the research questions can be formulated as follows:
- Q. What is the mapping between the topology of the physical space, as revealed through map annotation functionalities, and the semantic structure of the interaction space, generated by text-based communication?
- Sub-1. Does the time, semantic or social structure of the interaction space constitute a heuristic for pruning the space of asbtractions of the physical space? Do physical (map) distances relate to the social/time clustering of messages?
- Sub-2. Do users annotate the map in way that enables pattern matching mechanisms to induce the space-communication invariants? How large should the message database be (how many users / how much time) to be able to find patterns?
- Sub-3. Will a search engine based on space-interaction patterns support search processes that are less formal that keyword-based search process?
The main question is concerned with the integration of geographical, social and semantic elements into a model with the implicit aim to serve and improve the recovery process. Furthermore, subquestion 1 expands in this direction focussing on how the descriptors individuated by the main algorithm can be reported visually on the map for the same end results of the main question. On the technical side, the third subquestion focus on the idea that this discovery process is going to be finalised and tested through user experience observation.
Approach
The premises of the study will be tested on several group sizes, from small groups to communities.The idea is to give them this technology for a relatively long period of time, namely 6 months, during which some interventions are operated: different algorithms will be subsequently enabled in the systems, allowing the user to interact with different sets of functionalities. These functionalities will enable an improved search capabilities of the information contained in the system. The extended search engine will be based on geo-semantic pattern recognition. The user behavior and communication will be logged at all time. A comparison between the logs of groups without the extended functionalities and groups with these functionalities activated will constitute a proof technique towards the raised hypothesis.
An implementation of this algorithm will be applied, tested, and refined in a mobile geographical messaging system developed by our group (see figure 1), and in a Location Based Service implemented at EPFL scale, called ShoutSpace see http://craftsrv1.epfl.ch/research/shoutspace/ and figure 2).
Figure 1. STAMPS mobile client interface
Figure 2. ShoutSpace client interface
CRAFT is collaborating with Proboscis, UK, to exchange knowledge and technical data for research into spatial annotation and cognition. We are embarking upon a common analysis of the relationship between semantic and geographical dimensions of message distribution in space, based on the dataset provided by the 2003/04 Urban Tapestries field trials.
Kurze Beschreibung
STAMPS (auf deutsch Stempel) ist ein kleines Programm für Ihr Handy. Hat man es installiert, erscheint auf dem Display der Stadtplan des Orts, an dem man sich gerade befindet. Fügen Sie in den Stadtplan eine Nachricht (eine Art SMS) ein, können Ihre Bekannten diese Meldung auf ihrem Handy lesen. Schreiben Sie an die entsprechende Stelle: "Das ist meine Lieblingspizzeria!", werden einige Ihrer Empfehlung folgen und das Lokal testen. Alle Nachrichten, die in diesem System veröffentlicht werden, bringen anderen die Stadt näher: die Sportanlagen, die Kneipen, die Treffs. Je reichhaltiger diese Informationen werden, desto besser lernen Besucher die Stadt kennen. Sie können das System nach der nächsten Pizzeria fragen, und das System liefert alle Antworten, die es dazu hat. Ein kollektives Palaver über das Alltagsleben beginnt. (thanks to Reinhold Grether)
Short Description
STAMPS is a little program. It can run on your Mobile phone. Using this program you can see a map of the place where you are, visualised on the screen of your mobile. There, you can write a kind of SMS and attach it to the map so that other friends can see your message appearing on their map. You can write for instance: "this is my preferite pizzeria!", to offer advice to your buddies. All the messages left in the system say something about the city where you live: what are the sport locations, the place to eat, the meeting spots. After a while, we want to use all these information to help the users to navigate the city. You can ask the system, for instance: "where is a pizzeria near by?", and the system will search for other people's messages which refer to the term pizzeria to give you an advice.
Description Courte
Stamps signifie "System for TAgging Messages and Post-Inferential Semantics". Nous étudions la façon dont les utilisateurs font des inférences quand au message qu'il reçoivent. En particulier, nous sommes intéressés par la façon dont ces processus se produisent au niveau de la connaissance, pour être distingués du seul niveau linguistique. A terme, nous souhaitons construire un modèle formel des mécanismes de 'grounding' dans la communication géolocalisée.
Descrizione Breve
STAMPS (francobollo in italiano) è un piccolo programma che funziona su cellulari, la cui funzione principale è la visualizzazione della posizione dell'utilizzatore su una mappa sullo schermo del cellulare stesso. Da li è possibile scrivere dei messaggi tipo SMS che vengono condivisi con altri utenti attraverso la medesima mappa. Si possono scrivere messaggi del tipo: "questa è la mia pizzeria preferita!", allo scopo di consigliare i nostri amici. Tutti i messaggi archiviati, offrono preziose informazioni sulle risorse della città: quali sono i posti per fare sport, i posti dove mangiare, i posti per incontrarsi, sono solo una piccola parte delle possibilità offerte. Queste informazioni possono essere utilizzate per aiutare altre persone a muoversi attraverso la città. Ad esempio, si puo' chiedere al sistema: "dove si trova la pizzeria piu' vicina?", e il sistema cercherà nei messaggi degli altri utenti qualsiasi riferimento alla parola pizzeria per offrire un consiglio.
Publications
Cherubini, M. (2004). A collaborative ontology for ‘spatialised communication’. In Position paper for the workshop “Potential of Cognitive Semantics for Ontologies”, part of FOIS2004 , Torino, Italy. Available from:http://musil.uni-muenster.de/workshop2004/
Cherubini, M. (2005). Collaborative annotations of space in a mobile context: a computational model that integrates spatial information and communication. Thesis Proposal 2, CRAFT, Ecole Polytechnique Fédérale de Lausanne, EPFL, Ecublens, Station 1, CH-1015, Switzerland, 2005. Available from:http://www.i-cherubini.it/mauro/publications/
Cherubini, M. (2006). Phd annual report. Technical report, Ecole Polytechnique Fdrale de Lausanne, Ecoublens, Station 1, CH-1015 Lausanne, Switzerland, 2006. Available from:http://www.i-cherubini.it/mauro/publications/