Questions and Answers of Conceptual Structures In Practice

Modify the regular expression in Figure 7.6 so that it looks for a particular sentence pattern, for example, verbs that are followed by two prepositional phrases. [R]
Figure 10.9 shows the evolution of the line diagram used to display the concept lattice in the mail-sleuth project. Discuss step-by-step the changes that lead from the traditional line diagram of a
Can you identify other parts of speech that could be clustered using this method and determine suitable regular expressions for these part of speech? [R]
Write software that automates steps 2-5 from the workflow. (This could be done using the NLTK libraries or a scripting language.) [P,R,!]
Section 10.2.1.1 describes the idea of transitive closure on an attribute hierarchy namely asserting “a is-a b” and “b is-a c” will automatically infer that“a is-a c.” Characterize
What is SMTP? Find and describe the schema for SMTP. What extensions to SMTP are implemented in Microsoft Outlook that are not part of the de facto SMTP standard?
What is the difference between the term extent and contingent used in concept lattices? Give an example and develop a mathematical characterization of the terms.
Section 10.3.6 describes the idea of query-by-example in the context of image search. Research other popular techniques that implement query-byexample and discuss how these might be deployed to help
Regarding the design of a usability experiment, what types or task would you include in the usability test for an email client?
The last paragraph in the section on Roget’s Thesaurus discusses ways in which both Figure 7.7 and Figure 7.10 could be improved by adding more words, etc. Perform these improvements. With respect
Choose a lexical field of a verb and analyze the verbs in the field using RT neighborhood lattices, bilingual lattices and lattices for verb clusters.
Discuss how you might use the graphs (with instantiated values) to reason about the system in question.
In analogy to the example in Figure 7.5 using a dictionary that contains sufficient amount of detail, try to model the entries of a word with several senses or a lexical field of related words as a
In analogy to the example in Figure 7.4, can you model the relations for a word or set of words in WordNet (http://wordnet.princeton.edu/) in a similar manner using FCA? How can you derive a formal
Follow the instructions given in the workflow in Figure 7.8 for a word or set of words of your choice. Experiment with different words. Try the restricted neighborhood for some of the words.
Follow the instructions given in the workflow in Figure 7.11 for a word or set of words of your choice. Experiment with different words. Can you find any words where Steps 3 or 4 do not add any more
Choose a lexical field and analyze the words in the field using both RT neighborhood lattices and bilingual lattices.
Download a bilingual dictionary from the Web. Such dictionaries should be available for most major language pairs. You want to look for one that uses a format that is easy to parse (e.g., a tab
Neighborhood closure: implement an algorithm that calculates the neighborhood closure for any object or attribute of a formal context, i.e., the set of objects and attributes that belong to the same
Install the version of the Brown corpus from the NLTK distribution.Then follow the instructions in the workflow in Figure 7.6.
In Section 10.3.7 the idea of concept similarity and distance is defined. Research other ways that vector space models of information retrieval implement distance and similarity. Start with text
This chapter introduced a basic concept type hierarchy. Find a number of state-of-the-art research articles on online communities, Web services design, etc. Using these articles, refine and expand
The functionality mappings presented in this chapter are only very basic.Improve the definitions given, e.g., currently functionality requirements concern individual functions, but probably
Explain how the graphs are related to each other.
Import the files generated into Siena and generate diagrams. Discuss which diagrams are useful and which ones are not. Identify approaches to avoid some of the issues encountered.
Create a program/script that reads part of a file system and exports a number of different contexts in CXT or OAL format, including:• the size of the files found;• the owner of the files;• the
Create a ToscanaJ system based on this database using Elba. Try to include:• views for the individual objects;• reports for object sets;• a specific query for the labels;• some online help
Select one data source available for download and import it into a relational database.
Identify other data sources suitable for a ToscanaJ system. Discuss how they compare to the IMDB data.
Discuss how a CIS can be set up for a data source such as the Internet Movie Database (IMDB)18. Address the following questions:• Which parts are suitable as objects?• Which parts are suitable as
Choose a database or other source (e.g., a spreadsheet) whose data is familiar to you. Specify the source.
Show at least two instances of each graph populated with actual values from the database.
Draw graphs that represent the semantics of at least two tables or relations from the database using actors.
Define a partial homomorphism from an SG to an SG. How could this notion be used for approximate reasoning?
Select two cases of (online) collaborative communities, for example e-learning communities, gaming communities, or corporate communities. Using the (revised) concept type hierarchy and functionality
The example of a functionality matching process given in this chapter was that of implementing a functionality requirement. However, there are many more applications of such processes. One example
Much advanced Web services-related research is focusing on standards like the Universal Description, Discovery, and Integration (UDDI) protocol,15 Business Process Modeling Notation (BPMN),16 and
We have used the projection operation for the matching process. What roles could other core CG operations (e.g., maxJoin) play in developing new ways of functionality matching and design? How could
Amine is a powerful open source platform for knowledge system implementation,18 with a strong grounding in conceptual graphs. In particular its Prolog+CG module is very useful for CG representation
Is it possible to consider conjunctive relation-types in the same way as conjunctive concept types?
Let G and H be two SGs and π a homomorphism from G to H. What are the properties of two relation nodes r and r′in G such that π(r) = π(r′)?At what conditions can two relations nodes of an SG
Propose a mechanism for concept type definitions. Give a FOL semantics concerning your proposal. What are the questions to be solved concerning this concept type definition mechanism?
Create a ToscanaJ system based on multiples of these diagrams. Use ToscanaJ to create some nested line diagrams and discuss their usefulness for this application.