Lexical Database of Pakistani Regional Languages Chapter 5 Experimental Results
5.1 FRONTEND (USER WEB INTERFACE)
Web interface developed for this project is shown below. A text box is given where a user can write a Punjabi word and click ‘Search’ button. The word is searched in the database word-table and retrieved its id. The word-id is searched then in the asgn-word-sense-table and retrieved all the respective sense-ids. These sense-ids represent the unique senses in the sense-table from where the sense-text is retrieved. In this way, the given input word, all its senses and words corresponding to these senses are retrieved. For example when the word آ is searched the following output will be shown.

![]() |
![]() |
A user can enter the Punjabi words by using its own keyboard or using the on screen keyboard (OSK). When a Punjabi word is being searched, the result set consisted of word, grammatical category, sense and set of synonyms will be shown.

5.2 PROBLEMS DETECTED
5.2.1 Diacritics
Due to the presence of diacritical marks, it is very difficult to detect the exact word a user intended to search for. The simple solution is that the storage of all words with and without diacritics (aerabs). When a user enter a word, get the word free from diacritics and search the database and make results based on it. Some other strategy may also be applied to get the exact word for the user.
5.2.2 Dialects
Due to the presence of different accents and dialects of a language, it is very difficult to store all the variations of a word spoken by different diasporas of the same language.
Go To: Chapter-4 Go To: index