Lexical Database Software

James A. Landau JJJRLandau at AOL.COM
Mon May 17 22:33:42 UTC 2004


In a message dated  Sun, 16 May 2004 19:06:48 -0700,   Dave Wilton
<dave at WILTON.NET> writees
>
>  What I am looking for is much simpler. I am simply looking for a database
in
>  which to store citations for particular words or phrases. The electronic
>  equivalent of note cards.
>
>  This is a trivial use of a relational database. I was only hoping that one
>  would was already created, saving me several days of work.

THE FOLLOWING DATABASE SCHEMA IS PRESENTED WITH NO WARRANTIES, EXPRESS OR
IMPLIED

Where a name appears in two different tables, those tables are related on
that field.

Citation Master Table
     citation_ID    primary key
     date_citation_entered
     (bibliographic data fields TBD)

Citation Text Table
     citation_text_ID     primary key
     citation_ID
     word_whitespace_switch
     word
     root_word
     word_position_in_citation
     character_position_in_citation
     word length

Root Word Table
     word     primary key
     root_word
     ambiguity_flag
     compound_word_1
     compound_word_2
     date_last_updated

Phrase Historical Table
     phrase_ID    primary key
     date_last_searched
     number_of_words_in_phrase

Phrase Text Table
     phrase_text-_ID    primary key
     phrase_ID
     word
     word_position_in_phrase

Phrase Citation Table
     phrase_citation_ID     primary key
     citation_ID
     phrase_ID
     first_word_position_in_citation

The first three tables are updated automatically whenever a new citation is
added.  However, to be useful, the Root Word Table must be hand-edited.  The
last three tables are updated whenever a search is made on a phrase rather than
on a single word.

In the Root Word Table, the root word for "rings" and "rang" is "ring".
ambiguity_flag is set (manually) whenver a word can be either its own root word or
a derivative of another root word, e.g. "rung" is either its own root word
("rung of a ladder") or the past participle of "ring".  compound_word_1 and
compound_word_2 are set manually for a compound word which might be searched
either for its full text or for its component parts.

This schema will allow searches for two words or phrases "near" each other,
where "near" can be mesaured in either words or character positions.  Wildcard
searches are also possible and are limited by what ACCESS will allow.

It is ncessary to set an arbitrary limit on the number of words in a phrase.
Also as written a three-part compound word cannot be handled.

           - James A. Landau



More information about the Ads-l mailing list