Landmark:

landmark {
landmarkid: (13-digit code) AAAAABBBBBBBB (A - county code, B - serial number),
landmarkcode: (3-digit code) (classification code),
landmarkname: String,
landmarkaddr: String,
landmarkgeom: (geom)
}


Road:

road {
roadid: (13-digit code) AAAAABBBBBBBB (A - county code, B - serial number),
roadcode: (4-digit code),
roadaddr: String,
roadgeom: (geom)
}


Administrative Division:

admdiv {
admdivid: (9-digit code) AAAAABBBB (A - county code, B - serial number),
admdivcode: (10-digit code) AAAAABBCCC (A - county code, B - town code, C - village code),
admdivname: String,
admdivgeom: (geom)
}


http://www.iot.gov.tw/np.asp?ctNode=1079

Taiwan Review

Taiwan Review is an English periodical issued by GIO Taiwan.  The purpose of the periodical is to summarize the current status of art, society, and other events in Taiwan.  They have editorials, interviews, and articles with figures and tables.

They designed the ontology as follows: the main class is bibliography, which has sub-class “Taiwan Review” and sub-sub-class “Vol.# …”.  Articles in each volume are main instances in this work.  They designed other aspects as attributes and indexing factors for the articles, including authors, persons mentioned (classified by the category), location (“Taiwan” is taking out of “Asia” because there are more items under Taiwan than other places in Asia.  This is not logically sound, but may be more useful for users to navigate and query in practice.), and time span (They used decades before, but professor suggested to use events as separators of time span.)

Tea

Tea is an unique culture in East-Asia, but there are less focus on the knowledge organization about tea leaves, tea tools, and agriculture.  The main class is tea, and it has relationship with the one who raise the tea leaves, the place, the making process, and related tea tools to make the tea.  Because the student has the domain knowledge, this ontology is very well-organized and fit the application.

Music of Bunun

This work uses data and research results from other work to create the ontology.  The music is separated into songs and instruments, and it records who plays the music at what situation and location.  Some research results (domain knowledge) helps design the concepts and relationships.

Summary

Ontology is a tool to capture the concepts, relationship between concepts, and instances.  The basic relationship between terminology are broader term(BT), narrower term(NT), and related term(RT).  The design principle includes Clarity, Coherence, Extendibility, Minimal encoding bias, and Minimal ontological commitment. (Gruber 1993) There are several methodologies to design an ontology, and the components and steps proposed by Standford Protege team are domain, terminology, concept, relationships, and instances.

In determining the domain, several questions should be answered: What is the expert domain?  Why we need this knowledge organization?  What problems/questions should this KO answers?  Who will use and maintain the KO, and how?  Then we should search if there is similar ontologies, list the terminologies, define the classes/concepts and the relationship.

• III Taxi Project After discussion with Taiwan Taxi, we learned more about the current operation of taxi dispatching and the attitude of the company toward the hotspots. I think we all agree that the hotspot statistics and prediction is helpful for drivers to make decisions, but how to give instructions according to the prediction is a controversial thing.
• Paper Survey
  • SPIRIT Project The SPIRIT system wants to answer the queries which are in the form: , such as “restaurants near NTU.” Each geographical location is represented by a footprint. Applying different operations according to the rel to the footprint of where results in another footprint for the query [4]. It supports in, near, outside, east, south, west, and north, and it uses geo-ontology and domain ontology to derive the weights for adjusting the footprint generation. And to construct the ontologies from dierent databases, four matching mechanisms [3] are used to integrate the data. Footprint matching is the most strongest indication, but difficult to do due to the accuracy of the footprint definition in two datasets.
  • REXplorer Ballagas et al.[1] creates a gameplay for tourists to enjoy visiting the historical town with more interactions. It calculates optical flow from camera stills to estimate the movement of mobile devices, and checks whether players cast the right spell. It also maintains a travel blog on the server to track the progress, and the photos and videos players take on the way will be automatically tagged and attached to the blog entry.
  • Snap2Play Chin et al.[2] asks players to find a digital card in the real world, open the card and get a target photo, and match it with real scene. It uses GPS navigation for directing players, compass and accelerometer to detect the facing direction, and augmented reality techniques to immerse virtual objects to the real world. It uses SIFT as features and AdaBoost as classifier to match the photos.

What I plan to do.

• III Taxi Design geo-spatial ontology.
• Android
• AI Qualify Exam
• Paper Survey

References:

1. Rafael A. Ballagas, Sven G. Kratz, Jan Borchers, Eugen Yu, Steffen P. Walz, Claudia O. Fuhr, Ludger Hovestadt, and Martin Tann. Rexplorer: a mobile, pervasive spell-casting game for tourists. In CHI ’07: CHI ’07 extended abstracts on Human factors in computing systems, pages 1929–1934, New York, NY, USA, 2007. ACM.
2. Tat-Jun Chin, Yilun You, Ce’line Coutrix, Joo-Hwee Lim, Jean-Pierre Chevallet, and Laurence Nigay. Snap2play: A mixed-reality game based on scene identi¯cation. In MMM, pages 220–229, 2008.
3. Gaihua Fu, Christopher B. Jones, and Abdelmoty Alia I. Building a geographical ontology for intelligent spatial search on the web. In M. H. Hamza, editor, Databases and Applications, pages 167–172. IASTED/ACTA Press, 2005.
4. Gaihua Fu, Christopher B. Jones, and Abdelmoty Alia I. Ontology-based spatial query expansion in information retrieval. In On the Move to Meaningful Internet Systems 2005: CoopIS, DOA, and ODBASE, volume 3761 of Lecture Notes in Computer Science, pages 1466–1482. Springer Berlin / Heidelberg, 2005.

• III Proposal (Collaboration with Hsiao-wei, David, Yu-chin and Wan-rong.) We finished the III proposal about taxi hot spot prediction, and also revised it based on the review result from III.
• AI TA Work I graded MSRS task 2 and term project source code together with Tsu-yu, and created the archive CD for the department office.
• AAMAS Workshop Paper (Collaboration with Yuhana, Wan-rong and Joe.) I joined the discussion about paper outline and content, provided opinions to revise the problem definition, and edited/revised the chapter of introduction, background technology and related work. This paper editing is still on-going, and we expect to submit the paper by Feb. 8.
• Android Developer Challenge (Collaboration with David and Hsiao-wei.) We started discussion about the topic, and we decided to create an application combining the elements of city guidance and urban games. The idea is to provide users a city guidance service to travel in an area, using games as a motivation to have people enjoy. The system first designs a tour path and spots for the user considering the context information. When users arrive the spot, like REXplorer[3] and Snap2Play[5], a specific game which has some connections with the context will be given to the users. We have thinked about using ontology to model the time and location context, game attributes and user experience, and using reasoning to suggest the proper game for one spot and the next destination. However, to minimize the effort and meet the deadline, we will just create a prototype with all information fixed for idea demonstration purpose.
• Course Term Projects
• Paper Survey
  • Ontology SOUPA[4] is a well-known upper ontology for pervasive computing and context-aware systems. It defines (or modifies from other work) time, space, event, and some other core ontologies, and creates a context broker architecture (CoBrA) and two small scenarios to demonstrate the power of ontologies. It is an innovative work to create the ontologies to model context, and we can adjust their ontologies to meet our demand.
  • Geo-spatial Model Ye et al.[10] designs a more general space model to deal with both syntatic/semantic label and geographic position, both resolute referencing and relative referencing, and both containment and connection properties. I have to say that they do consider more aspects of describing geo-spatial information than other work, and this model seems more powerful to deal with different kinds of location representations and to provide services.
  • Navigation Systems To write the related work of the AAMAS Workshop paper, I quickly skimmed through several papers[1, 2, 8, 9].

What I plan to do.

• AAMAS Workshop Paper Finish editing the paper and submit before the due date.
• CA Group Think of one location-aware service, and create a small ontology for the scenario.
• Android Developer Challenge Pick one city tour itinerary, collect relevant photos and descriptions, and design games for introducing the spot. These will become part of the demonstration content of our prototype.
• Paper Survey
  • Spatial Queries Using geographical ontology for spatial queries[6, 7] is one part of SPIRIT Project. I think spatial queries and spatial indexing problems are very alike, and I hope I can find something from their work.
  • Mobile Urban Games REXplorer[3] and Snap2Play[5] introduce two kind of mobile urban games, and I hope I can find some more examples or ideas from their related work.

References:

1. Gregory D. Abowd, Christopher G. Atkeson, Jason Hong, Sue Long, Rob Kooper, and Mike Pinkerton. Cyberguide: a mobile context-aware tour guide. Wireless Networks, 3(5):421–433, October 1997.
2. Christos Anagnostopoulos, Vassileios Tsetsos, Panayotis Kikiras, and Stathes P. Hadjiefthymiades. OntoNav: A semantic indoor navigation system. In Proceeding of the first International Workshop on Managing Context Information in Mobile and Pervasive Environments, 2005.
3. Rafael A. Ballagas, Sven G. Kratz, Jan Borchers, Eugen Yu, Steffen P. Walz, Claudia O. Fuhr, Ludger Hovestadt, and Martin Tann. Rexplorer: a mobile, pervasive spell-casting game for tourists. In CHI ’07: CHI ’07 extended abstracts on Human factors in computing systems, pages 1929–1934, New York, NY, USA, 2007. ACM.
4. Harry Chen, Filip Perich, Tim Finin, and Anupam Joshi. SOUPA: Standard ontology for ubiquitous and pervasive applications. In First Annual International Conference on Mobile and Ubiquitous Systems: Networking and Services (MobiQuitous’04), pages 258–267, August 2004.
5. Tat-Jun Chin, Yilun You, Ce’line Coutrix, Joo-Hwee Lim, Jean-Pierre Chevallet, and Laurence Nigay. Snap2play: A mixed-reality game based on scene identi¯cation. In MMM, pages 220–229, 2008.
6. Gaihua Fu, Christopher B. Jones, and Abdelmoty Alia I. Building a geographical ontology for intelligent spatial search on the web. In M. H. Hamza, editor, Databases and Applications, pages 167–172. IASTED/ACTA Press, 2005.
7. Gaihua Fu, Christopher B. Jones, and Abdelmoty Alia I. Ontology-based spatial query expansion in information retrieval. In Robert Meersman, Zahir Tari, Mohand-Said Hacid, John Mylopoulos, Barbara Pernici, Ä Ozalp Babaoglu, Hans-Arno Jacobssen, Joseph P. Loyall, Michael Kifer, and Stefano Spaccapietra, editors, On the Move to Meaningful Internet Systems 2005: CoopIS, DOA, and ODBASE, volume 3761 of Lecture Notes in Computer Science, pages 1466–1482. Springer Berlin / Heidelberg, 2005.
8. Alan L. Liu, Harlan Hile, Henry Kautz, Gaetano Borriello, Pat A. Brown, Mark Harniss, and Kurt Johnson. Indoor wayfinding: developing a functional interface for individuals with cognitive impairments. In Assets ’06: Proceedings of the 8th international ACM SIGACCESS conference on Computers and accessibility, pages 95–102, Portland, Oregon, USA, 2006. ACM Press.
9. Fernando Lyardet, Jan Grimmer, and Max Muhlhauser. CoINS: Context sensitive indoor navigation system. In Proceeding of the Eight IEEE International Symposium on Multimedia (ISM’06), 2006.
10. Juan Ye, Lorcan Coyle, Simon Dobson, and Paddy Nixon Nixon. A unified semantics space model. In Location- and Context-Awareness, volume 4718 of Lecture Notes in Computer Science, pages 103–120, 2007.

2. Plug in the bluetooth receiver, and go through the settings.

1. Survey current ontology specs on time, location, and event.
   • W3C
   • SOUPA project
   • SPIRIT project
   • GeoNames project
   • etc.
2. Survey about geo-spatial indexing methodology.
   • Li W., Yang C., and Zhou B., 2007. Geospatial Information Retrieval , in Encyclopedia of GIS , Bin Xiong and Shashi Shekehar (Eds), Springer-Verlag, (in press)
   • Vonoroi Diagram
   • etc
3. Finish the III proposal about taxi hot spot prediction before 1/10.
4. Create the first version of location ontology for general usage.
5. Create the first version of time ontology for general usage.
6. Create the first version of event ontology for general usage.
7. Create a small android application with MapView included.

I decide to try wordpress from the beginning of 2008, and hope I will love it.

Hanwen