The invention relates to the technical field of bidding and
purchasing information recommendation, in particular to a bidding and
purchasing information recommendation
system and recommendation methodbased on a
knowledge graph. The recommendation method comprises the steps: S1, carrying out the manual collection and classification of bidding and
purchasing industry knowledge, building a mode graph of a
knowledge base, positioning the collection direction of a data crawler, and
crawling data; S2, cleaning and arranging the collected data; S3, establishing an indication graph, reasoning announcements by using the
knowledge graph, and learning according to knowledge contained in the announcements; and S4, intelligently recommending and pushing announcements to the user by using the deployedand operated
knowledge graph. For the bidding and purchasing information recommendation
system and recommendation method, structured or
unstructured data such as domain classification and encyclopediacontent are crawled by adopting distributed crawlers, and a knowledge graph
library is opened to obtain data; bidding and purchasing information is intelligently recommended, and similar projects arefound out through project announcement titles and contents; similar users are obtained through the bidder portrait, then projects which the similar users are interested in are obtained, and projectswhich the bidder may be interested in are recommended to the bidder.