Service combination package recommendation system and method based on text mining

Abstract

The invention provides a service combination package recommendation system and method based on text mining. The method comprises the steps that 1, historical service combination data is collected to construct a data set, and historical service combinations and historical service combination function descriptions in the data set are constructed into a service database; 2, the historical serve combination function descriptions in the data set are partitioned into a plurality of statement blocks according to a rhetorical relation; 3, an intersentential relationship tree is constructed for each service combination to obtain relationships among the different statement blocks; 4, relationships among different services are acquired; 5, the relationships among the services are counted to generate a relationship matrix; 6, a user need is divided into statement blocks, and the previous K services are recommended to each statement block, wherein K is a positive integer; 7, semantic matching is conducted on the services in each statement block by combining the relationship matrix, and a service package recommended to a user is obtained. According to the system and method, the user can be helped to better build a model, the service combination package is recommended to the user, and then the overall effect of service recommendation is improved.

Description

technical field [0001] The invention relates to the technical field of Internet service retrieval, in particular to a system and method for recommending service packages based on text mining. Background technique [0002] With the development of the service computing field, the implementation and carrying forms of services have become more and more abundant after a long period of development, showing a trend of diversification. Providing customized services for users with different needs is gradually attracting attention. Traditional service providers are unable to provide customized services for each specific need, therefore, service portfolio becomes a solution to provide personalized services. For example, if a user needs to find a suitable restaurant to eat at their location, they need to list the local restaurants through the map service, and then obtain restaurant information on the evaluation service. If the two services are combined, it can provide personalized s s...

AI Technical Summary

Problems solved by technology

Currently, users search for services through Web service search engines, but keyword-based matching often fails to find truly suitable services for users.

On the one hand, existing service-composition-oriented recommendations have a cold-start problem, that is, they cannot recommend new services; on the other hand, they do not consider the mutual cooperation of services, that is, they cannot recommend services with complementary functions.

Images