Project Detail

Description

60 - 70% in the project
see attached files for full details.

Develop a standalone Java application which works as a single-user application for displaying concordance of a given corpus. This application is expected to work with texts that are in UTF8 encoding. The given texts are chosen from various British novels and poetry. All texts can be found under the folder corpus, a zip archive.

Your task is to use J2SDK 6 to design and implement the intended concordance application that will work with the given text files.

Folder corpus contains seven text files which amount to 29,361 words:

1. a christmas carol chapter1.txt
2. a christmas carol chapter4.txt
3. emma chapter1.txt
4. emma chapter2.txt
5. pride and prejudice chapter3.txt
6. pride and prejudice chapter42.txt
7. spirits in bondage.txt

Each of the first six pieces of texts is a chapter from a famous novel. Each paragraph in
a text is separated by a blank line. The last text spirits in bondage.txt is made up of
verses. Each line in a verse is listed on a separate line.

>>>Functional Requirements
The intended application will support the display of concordance in the KWIC view (cf. Figure 1). The context is defined as 10 words to the left and 10 words to the right of the required word.

1. At the start of the application, all text files in the given data folder (i.e. corpus) will be
read into the system. To facilitate the lookup, for each word in the text file, an index of
the word and its contexts will also be built.

Hint: A word is defined as a sequence of characters delimited by space. These characters include:
_ letters in the English alphabet: a, b, c, . . . , y, z
_ apostrophes: ’
_ hyphens: -
For example, each of the followings is considered to be a single word:
_ Scrooge _ Bennet’s _ I’ll _ new-born _ brother-inlaw
However, each of ways--with, London--his and altogether--Mr will be considered as two
words, i.e. ways and with, London and his, altogether and Mr, respectively.
Each entry in the word-context index should unique. Hence, if the corpus contains the same word more
than once, the word will appear in the index once, i.e. as a single entry.


2. A user can look up a word and the system will display its concordance in the KWIC
view.

3. A user can exit the application.

>>> Non-functional Requirements
1. The processing time of each search operation must be short.
2. A user will interact with the system through a Text-based User Interface (TUI).
3. The display of results for each query must be clear and easy to understand, with the
search word for each result aligned neatly along the same column.
4. The application must be robust and display appropriate messages should any run time errors (including no item found) occur.
5. The application must include the use of appropriate confirm windows for obtaining
user confirmations.


Additional files submitted:
Spec.doc
corpus.zip