Research On Key Algorithm Of Trip Recommendation Based On Multi-constrain And Multi-objective

For the past few years, with the vigorous development of the tourism industry and tourism consumption characteristics of the tourists have also become more rational and personalized, How to find attractive tourism products as tourists, travel agencies and tourist important issues of common concern to the site. In tourism website such as lvping website evaluation of network, a large number of tourists has released mass tourism related information. However, these information are often very scattered, the lack of structural features and common visitors to set according to their own tourism planning. Therefore, trip recommendation research gradually got the attention of many scholars. However, most existing methods only consider the constraints in a certain respect of the user, but most users travel, often by cost, time, transportation and other various constraints, based on a single constraint recommendation result is difficult to meet the real needs of users.In this paper, based on the Internet to provide huge amounts of data, we research route recommendation methods based on multiple attribute evaluation mechanism and more constrained multi-objective tourist. First of all, the development of the social network platform for travel information service provides a rich source of data and sharing platform, therefore, through mining the various information of attractions from social networking sites, such as scenic spots open time, ticket information and attractions, such as the GPS coordinates, etc., we can get more comprehensive and reasonable assessment. This paper proposes an attraction evaluation mechanism based on multiple attribute, it considers a variety of properties of scenic spots, comprehensive score of attractions.Secondly, under the condition of multiple constraints, tourist route recommendation algorithm complexity is very high, we research a kind of k-greedy recommendation algorithm under multiple constraints. Its key idea is, based on a given user constraints, using the greedy algorithm to select the k before the overall score higher path, then, through the diversity and efficiency of the comprehensive evaluation path, to choose a relatively optimal path as the recommended result.Based on the data sets mined from lvping website, we has set up a prototype system for the trip recommendation. The experimental results show that, first of all, through the fusion of attribute, the resort comprehensive evaluation mechanism makes reasonable and comprehensive evaluation, the recommend laid a solid foundation for the subsequent route recommendation. Second, compared with the existing recommendation algorithm of Trip-Mine+ algorithm compared with better effect, k-greedy recommendation algorithm can satisfy the user’s various constraints and objectives, its recommendation has better rationality and diversity, and this algorithm has low time complexity in run time.