Paper Title:Understanding as We Roam
O'Sullivan, D.; Wade, V.; Lewis, D.;
Internet Computing, IEEE
Volume 11, Issue 2, March-April 2007 Page(s):26 - 33
Three Critical Questions
Monday
Group 1:
Member Name: Abilash.Amarthaluri,
Bharadwaz.Somavarapu.
Critical Questions:
1. As the author specifies in the paper, the user who wants to use the applications remotely must have to download the related file prior to the travel, Dynamism is always considered to be a problem in roaming applications. Does the user require downloading that file every time he plans a journey? Reusability issue is not discussed in the paper.
2. Abstraction is another problem needed to be considered. If a user goes into a remote location, there is no need for him to know how the process is implemented. i.e., the user must be given such level of abstraction so that he might not want to know how his own applications interact with the local computing resources.
3. Locale specific problems are also needed to be considered, as all the applications need not be based on a specific locale. There should be Locale independent applications shared among the network providers or there should be efficient mapping tools which can map the locale problems.
4. As there are no proper standards and “commonly agreed ontology mapping life cycle”, there is Interoperability problem.
Group 2:
Member Name:
Group 3:
Member Name: Sunil Kakaraparthi, Yaswanth Kantamaneni
1. Author relies on the decide to match kind of activity and analyzes the corresponding data to map the ontologies. How this matching is done when there are a number of matches found and on which terms and conditions the decision occurs?
2. Author tells in the paper that the applications are not used frequently but they are used based on the another application. In case if an application should depend on another application how the security of the applications is resolved and handled?
3. There are problems with the “HP -Jena-toolkit” like the process and the software is more vulnerable and it is not a perfect standard. How about any of the other toolkits which solve the above mentioned problems?
4. In the candidate with the multiple mapping what’s the scenario behind this process? How will this kind of mapping would effect in the expert and the individual use case?
Group 4:
Member Name:Nikhilesh Katakam, Ramya Devabhakthuni
1. The paper stated that one “candidate match” can have “two or more mappings”. This indicates that there will be many choices for mapping. Thus more human effort is required and how will the user identify the best mapping?
2. The number of “web services” and the “service providers” are increasing day-by-day. If this is the case, how the mapping will be done and how the interoperability be achieved?
3. The paper stated that there should be common way for the developers to indicate the “results of Matching Algorithms and Systems”. But for each specific domain there would be different mappings and different specifications. How this agreement will be achieved and how the solutions are provided to mismatch mappings and conflicts?
Group 5:
Member Name: Lokesh Reddy Gokul
• The author mentions the need for the mapping process to support choice of activities for an organization in it’s implementation. Such a choice to choose among the manual, semi automated, automated options available may sometimes be not the ideal case for the realization of semantic web , because in an ideal semantic web environment user interaction is to be minimum possible. So what would be an ideal design where most of the processing is automatic and the decision making for an organization is not a task to be taken care of?
• In the context of Candidate matches to Potential Mappings table, the author gives an example of the keyword Restaurant in various contexts and this example prompts a question that is the matching still a text-matching based approach? What are the alternative approaches possible to achieve a better mapping between ontologies?
• When the author discusses the dynamic semantic interoperability issues in various contexts that he experiments with his process model, he states the use of different techniques for different models. So this brings us to the issue that we need a uniform model for various models in case we need to perform ontology mapping on a run for user when he is roaming. So what are the issues in achieving such a uniform model?
Group 6:
Member Name: Palla Sangram
1) Can a platform which supports both software and human interpretable form be developed?
2) Author in his paper says that “Roaming devices will download the mapping either prior to the travel or when he comes in contact for the first time”. Then how can the dynamic requests are handled?
3) Author mentioned that developers will not use the same semantic models on web signatures. But if they do not follow the same signature, how can outside people who are outside the domain understand the concept?
Group 7:
Member Name: Kishore Kumar Mannava & Mohana Siri Kamineni
1. The “author” presents that several other “challenges” are to implemented for the “benefit of roaming”. Are any tools implemented to overcome the “challenges” and what are the consequences after satisfying these constraints.?
2. The paper propose that the problem with ‘roaming users” is “achieving interoperation” and “discovering local web servers”. But how this ‘interoperation” is achieved at the “semantic level” and “syntactic level”?
3. The paper states that “even in the same organization” there are many different “use cases” for “ontology mapping” and gives that the design of such a system is “nontrivial”. Then why to use these ‘mappings” for “roaming users”?
Group 8:
Member Name:Durga Maheswari Muppalla
1) How “mapping is achieved” between the “local service providers” and the “roaming users” when there are lot of “complex correspondences” where a property in local ontology is equivalent to the other property in the “roaming users” ontology and how the “mapping is achieved” when there is high ““diversity”” in the services.
2) How the “service standards” is maintained between the “local providers” and the “roaming users” when there are many issues like the “interoperability” and the “diversity in services”.
3) As “no two organizations” are similar and moreover different conditions and different “use cases” may exist in the “same organization” how the mapping can be done in the large organizations with many “roaming users” as the requirements keep on changing.
Wednesday
Group 1:
Member Name:pelluri,lattupalli,voruganti
When talking about decide to match process author said that if the system makes a positive decision then it goes to the next step. But how can a system decide whether it is a optimum decision or not? User is the one who should decide it.
When we consider realistic approach, it may become a cumbersome task to locate, map and interoperate different ontologies. Because real world has many applications so as the number of ontologies, it has totally open environment. Is it not difficult to embed all these techniques in a cost effective way in mobile devices?
Author suggested in using a common language for mapping. But different organizations may have different constraints for their business needs. How can all of them bind to one language? I think it is hard to make all the ontology developers to use same language.
Group 2:
Member Name: Bobbili Shrinath, Addagalla Satyanaga Manoj Kumar, Gopinath, Sreejith
The ontology mapping considers mobile application and local service environment. When talking about mobile applications the geographical constraints need to be considered. How will the mapping be done for a lethargic geographic surrounding?
There can be separate mapping technology for each local service provider. How will the mapping techniques vary for respective service providers?
The author narrates four techniques for mapping. Will these techniques overcome the interoperability of semantic web because the mapping is done based on unified modeling language? There can be better modeling languages than UML.
Group 3:
Member Name: Swati Thorve
1> Ontology mapping is one of the step for making roaming SW reality. However there are other factors also need to be considered like legal matters while accessing ontologies, ontologies availability, is it only for private use? Local and SW users applications interoperability etc. We need to consider all such factors for roaming SW. What will be the process cycle after taking into consideration all these steps ?
2>Ontology mapping process mentioned by author is based on linear development means following one step after other step. What if the local ontologies that are being mapped to roaming users ontologies are changed, they become unsuitable for mapping, or due to change their mapping will now become more costlier? How can we incorporate these ontology changes during process in the authors OSIN model?
3>While deciding on standard ontology mapping process, we should also think about how can we bring standardization in different sub steps involved in the process like for mapping phase, its not enough for everybody to follow the same phase but also they should follow it using same method. Process standardization is decided, in same way how we can we implement methods standardization at the same time making it generic for every bodies use?
Group 4:
Member Name:Shaiv, karuna priya rameshwaram, anusha vunnam
1) The ontology mapping requires using many ontologies and mapping tools. Will this maximize the costs required by the organization in building webservices and related ontologies?
2) What type of common standards have to be followed as there are many ontology mappings that follow different set of mapping syntactic rules.
3) The author has not clearly explained the approach in automating the mapping phase?
Group 5:
Member Name: Rahul Mootha,Rahul Reddy
• 1. The author says the main purpose of ontology is to let the system and users communicate and share an understanding of any domain. But with different approaches by different people how is it possible to make ontology meet the needs of all the users?
2. What are the technologies that are being used for manipulating the xml models, the paper discusses about java tools, but xml being a highly interoperable one are any other tools being used currently?
3. The web services itself is a upcoming field, it still has a few issues to be dealt which like security, ontology mapping similarly is an application which has recently come to the market, how feasible is it that the conjunction of these two would be implemented with success, what can be the estimated time?
4. What is the strategy that the ongoing research which will helps us know people’s ontology’s with internet channels and not through the explicit engineering?
Group 6:
Member Name:
Group 7:
Member Name: Goyal, Saurabh, De Morais Andrade,Pablo and Boda,Vamshidhar Reddy
• The first question regarding ontology mapping and the process proposed is: how much work not automated will be necessary in order to make the mapping of ontologies?
• The second question, still regarding the human work versus automate work is: who will have to do this mapping? And the new ontologies that will be created everyday? Is there any place for them in this process, are they gonna be mapped “on the fly” or they will have to be mapped later?
• The last question regarding the paper is regarding the use of the process. How far is this process been used? Where has it been applied? Is there any company that experienced this process before? Is it proved to be implementable for companies or so far only in a research environment?
Group 8:
Member Name:Brugu Kumar Bhargava
Author says regarding the execution case of mapping process that mapping info must be available in both human and software interpretable form. Can we develop a platform which can support both at the same time ?
2) Author says that "It is unlikely that service developers will use the same semantic models in their web signatures" If the developers does not follow that signatures how can people outside the domain can understand the issue ?
3) Author says that a framework approach can help company situate the ontology mapping process but every organization has its own way of framework to compete in the market. Both the statements contradict. How can we achieve the global acceptance?
Group 9:
Member Name: Satish Bhat, Holly Vo
1. Ontologies and ontology mapping are new in terms of representing roaming device context. How can it be made easy to understand and user friendly to increase is acceptability? Current web technologies are fairly good enough in terms of their service response and QoS.
2. Ontologies are either downloaded for service provider site or added manually. But there may be need to do mapping on mobile user devices. The capability of these devices is limited in terms of space and processing. Will this hamper the mapping process?
3. Can ontologies be applied to time critical application in this period? Ontologies query is currently time consuming. Ontology mapping is complex process and currently very slow process.
Group 10:
Member Name: Sunae Shin, Hyungbae Park
1) They suggested techniques help users analyze ontology quality or potential difficulties in mapping particular ontologies. However, they didn’t mention why it would be better than the systems that use simple lexical matching techniques. Intuitively, their techniques seem to work better than old techniques and actually they showed the better performances of the OISIN however they missed the reasons for the better performances.
2) They introduced the management phase for the ontology-mapping life cycle however they didn’t mention anything about this in the implementation. Actually, it has not been developed yet in their implementation. They need to mention the reason why they didn’t contain the management phase in their implementation.
3) The OISIN process is supported by eight different applications. However, there is no explanation about these applications. In the OISIN process, there are more than eight different applications. Thus, they need to specify which one is included in the eight applications.
4) Having layers (phases) in their OISIN process is good idea. These layerings help reduce the loads which could be burst to the modules as well facilitate the addition or substitution of applications.
5) They mentioned mappings can cater for language differences during roaming. However, they didn’t mention anything how the mappings can do it.