The Design And Development Of Fire-fighting Equipments Query System

It is necessary to query the Fire-fighting equipments before purchasing. Recently, there isa wide range of types of Fire-fighting equipments, and there are also many manufacturers toproduct the same type of Fire-fighting equipments. In order to capture the Fire-fightingequipment information that may be of interest, the query user submitted is often not veryrestrictive. However, this query style may lead to the problem of too many query answerresults which referred to the information overload. In reality, the requirements of firedepartment may be different during the different period, and thus the Fire-fighting equipmentsquery results should be categorization according to the multi-types of requirements of firedepartment to deal with the information overload problem.In order to deal with the information overload problem of Fire-fighting equipments query,this thesis proposes a Fire-fighting equipments query results categorization approach based onthe multi-types of requirements. The processing procedure of the approach includes twophases, the off-line pre-processing phase and the on-line query processing phase. In theoff-line pre-processing phase, the original data of Fire-fighting equipments should bepre-processed and the semantic similarity between a pair of different Fire-fighting equipmentscan be estimated by using the coupling relationships of them. And then, the Fire-fightingequipments can be clustered according to their semantic similarity. In the on-line queryprocessing phase, based on the cluster of Fire-fighting equipments generated in the off-linepre-processing phase, a leveled categorization tree with labels can be constructed dynamicallyover the query results set by using the modified decision tree categorization algorithm. Thebranch label of the tree makes user can easily select and locate the Fire-fighting equipmentinformation he/she needs.Finally, the Fire-fighting equipments query results automated categorization approach isapplied into the Fire-fighting purchasing system. The realistic application demonstrates thatthe approach proposed in the thesis can satisfy the user multi-types requirements, and has thegood categorization effectiveness and low searching cost as well, which means the reasonableand practicability of the approach.