Example Queries
This section shows some examples of the different types of queries.
Agent or Category Query
You can add agent (or category) weights to terms in your query. The IDOL Agentstore component returns documents that contain highly weighted terms with a higher relevance than documents that contain only terms with lower weights. Use the TermGetInfo
action to find out the weight of the term. Specify weights as a number in brackets next to the term. The maximum weight is 4095
, but OpenText recommends that you use a maximum weight of 511
.
For example:
http://localhost:9050/action=Query&Text=Cat[50] OR Dog
This query returns documents that contain the term cat or the term dog. If the term dog has a higher weight than 50
in Agentstore, documents that contain the term dog return with higher relevance than documents that contain only the term cat. If the term dog has a lower weight than 50
, documents that contain the term cat return with higher relevance.
http://localhost:9050/action=Query&Text=Cat[30] OR Dog[10]
This query returns documents that contain the term cat or the term dog. Documents that contain only the term cat return with three times the relevance than documents that contain only the term dog.
You can apply term weights to phrases. For example:
http://localhost:9050/action=Query&Text="cats and dogs"[100]
This query returns documents that contain the phrase "cats and dogs"
and applies a weight of 100
to this phrase.
You can apply weights to terms in parentheses. In this case, the weight applies to all terms within the parentheses. For example:
http://localhost:9050/action=Query&Text=(cat OR dog)[100] AND (fish OR dolphin)[150]
This query returns documents that contain at least one of the terms cat or dog, and at least one of the terms fish or dolphin. It assigns the terms cat and dog a weight of 100
, and it assigns the terms fish and dolphin a weight of 150
.
You can use multipliers to multiply the original term weight. Enter a term as *N
, where N
is the amount to multiply by. You can use up to two decimal places. For example:
http://localhost:9050/action=Query&Text=cat[*2.25] OR dog[*0.5]
This query returns documents that contain the term cat or the term dog. The term cat has 2.25
times its normal relevance. The term dog has 0.5
times its normal relevance.
You can also use any combination of these examples to apply term weights to complex phrases or expressions. For example:
http://localhost:9050/action=Query&Text="cats and dogs"[*3] OR (fish[100]+dolphin[150])[*1.5]
This query applies 3
times the normal weight to the phrase "cats and dogs"
. It assigns a weight of 100
to the term fish and a weight of 150
to the term dolphin. It then multiplies the weights for fish and dolphin by 1.5
.
Profile Query
http://localhost:9050/action=Query&Text= CHAMPIONLEAGU~[551] EVERTON~[407] BAYERN~[402] UEFA~[391] PREMIERSHIP~[383] FIFA~[257] STRIKER~[226] WORLDCUP~[215] EURO~[124] SOCCER~[114] CUP~[66]
This action sends a profile query to the IDOL Agentstore component. The query contains the terms that the profile training contains, and the weight of each of the terms. Agentstore can return agents, profiles, categories, or documents that conceptually match the terms of the query.
Text Query
http://localhost:9010/action=Query&Text=Gene analysis discovered methods to determine the exact sequence of nucleotides that compose a specific gene.
This action sends a text query to the IDOL Content component. Content can return agents, profiles, categories, or documents that conceptually match the query text.
Suggest Query
http://localhost:9010/action=Suggest&ID=10
This action sends a Suggest
query to the IDOL Content component. Content can return agents, profiles, categories, or documents that conceptually match the specified document (that is, the document with the ID 10
).
SuggestOnText Query
http://localhost:9010/action=SuggestOnText&Text=Gene analysis discovered methods to determine the exact sequence of nucleotides that compose a specific gene
This action sends a SuggestOnText
query to IDOL Content component. Content can return agents, profiles, categories, or documents that conceptually match the terms with the highest weighting in the query text.