Research On Knowledge Discovery And Recommendation-Based On Text Mining Of Travel Notes

Nowadays,many people tend to get tourist information from the online tourism platform to make their travel plans closer to their own needs and preferences.In the meantime,historical tourists share and summarize travel experiences and write lots of travel notes on the online tourism platform.These travel notes are of great value to potential tourists.This paper takes the travel notes data of trips to Yunnan as an example,which are extracted from Mafengwo.Using text mining technology,combined with frequency analysis,co-occurrences analysis,and network analysis,the paper shows the tourist attractions and routes for trips to Yunnan according to knowledge discovery and aggregation on travel notes.Meanwhile the knowledge aggregation map for characteristics of the tourist attractions and travelers'experience are drawn out.Moreover,based on the results of knowledge discovery,the labels for core topics in travel notes are formed.The paper computes travel notes'degree of description and loyalty in accordance with corresponding labels to get representative notes.Then,the selected results are used for information recommendation.The purpose of this study is to propose an effective strategy to extract and use potential knowledge in travel notes to serve information ordering and recommendation for online tourism platforms.At first,the paper explains the reasons for the selection of sample data,and the steps and methods of obtaining the experimental data are introduced in detail.Then,the paper makes a simple statistics on the external attributes of travel notes,including the features of user data distribution,and analyzes the feature of travel notes from the whole.After that it expounds the ideas and algorithm of knowledge discovery and aggregation of travel notes and tourism recommendation through representative travel notes.Then the experimental results and related analysis are given with sample data,and then the test and discussion are carried out.From the experimental results,the mining algorithm in this paper can discover real hot spots and key tourism demands of tourists in the previous tourism experience,and realize hot tour routes for recommendation.At the same time,Through the comparative analysis of our results with tourism information about Yunnan on Mafengwo,the paper puts forward the service advice on the online tourism platform which puts emphasis on the relationship between information and ensures the key information having great usefulness and timeliness.Besides,travel notes with high similarity have been distinguished and clustered through computing travel notes' degree of description and loyalty in accordance with corresponding labels to get representative notes.It has guiding meaning and practical significance for smaller granular recommendation of tourism information.