Need help in using ctx_doc.themes()

Question

Hi ,
I am new to oracle text , i have a table which has a 'e_id' column(primary key) and text data in one of the column which is been context indexed, now i want to find the frequently appearing words in that column, so i am using the ctx_doc.themes() procedure to achieve it.
I have created a result table 'ctx_themes' which has query_id,themes and weight columns which will get populated after executing the ctx_doc.themes() proc, but i am facing the below error after executing the ctx_doc.themes() proc, so please help me in fixing the issue.
begin
ctx_doc.set_key_type('PRIMARY_KEY');
end;
begin
ctx_doc.themes('text_data_index','e_id','CTX_THEMES',1,full_themes => FALSE,num_themes=>20);
end;
Error starting at line 1 in command:
begin
ctx_doc.themes('text_data_index','e_id','CTX_THEMES',1,full_themes => FALSE,num_themes=>20);
end;
Error report:
ORA-20000: Oracle Text error:
DRG-11445: rowid value is invalid: e_id
ORA-06512: at "CTXSYS.DRUE", line 160
ORA-06512: at "CTXSYS.CTX_DOC", line 210
ORA-06512: at line 2
20000. 00000 - "%s"
*Cause: The stored procedure 'raise_application_error'
was called which causes this error to be generated.
*Action: Correct the problem as described in the error message or contact
the application administrator or DBA for more information.
Thank You.

Answer 1

token_type = 9 is a stem, and the token count is the number of documents in which words with this stem occur.
In your case I guess it is coincidence that the count for the normal token "work" is the same as for the stem "work".
See this SQL*Plus test script:
drop table t;
create table t (text varchar2(2000));
insert into t values ('i am working today');
insert into t values ('i worked yesterday too');
insert into t values ('i will be working, working, working tomorrow');
exec ctx_ddl.drop_preference('mylexer')
exec ctx_ddl.create_preference('mylexer', 'basic_lexer')
exec ctx_ddl.set_attribute('mylexer', 'index_stems', 'ENGLISH')
create index ti on t(text) indextype is ctxsys.context
parameters ('lexer mylexer');
column token_text format a30
select token_text, token_type, token_count from dr$ti$i;Output is
TOKEN_TEXT                 TOKEN_TYPE TOKEN_COUNT
AM                         0         1
I                         0         3
TODAY                         0         1
TOMORROW                    0         1
TOO                         0         1
WORK                         9         3
WORKED                         0         1
WORKING                     0         2
YESTERDAY                    0         1