With the widespread application of Internet technology, the tourism industry has been developing rapidly. It has become the mainstream way to plan trips by searching for travel information. However, the problem of information overload in the tourism field has become increasingly prominent, making it difficult for tourists to find information that matches their needs. Existing travel websites only provide a massive amount of tourism information, including travel product recommendations, travel guides, and itinerary plans. There is a lack of services for self-planned travel routes on the market, which cannot meet the demand for personalized travel itineraries. Therefore, studying the problem of travel route planning has practical significance and commercial application value. Currently, the rise of mobile internet and big data has driven the booming development of online reviews for tourist attractions, accumulating a large amount of textual information. This paper proposes a method for travel route planning and navigation based on big data and neural networks. Specifically, this paper constructs a neural network for sentiment analysis of tourist attraction reviews, which can infer the satisfaction of the public with tourist attractions. The model extracts attribute words from the comments and filters and clusters the extracted candidate attribute words to obtain the basic attributes of the comments. In addition, this paper proposes a deep-learning model based on multi-head self-attention and gated convolutional neural networks to analyze the comments. After obtaining the analysis results, travel routes can be planned and navigated based on the ratings.