With the rapid development of the Internet,great changes have taken place in the way and speed of news transmission.The replicability,rapid dissemination and massive information of Internet news have brought challenges to the authenticity and credibility of the news.The study proposes an Internet news backtracking method based on blockchain technology,combines text clustering and deep learning technology,and quickly traces the origin and evolution of news with blockchain distributed storage and immutable characteristics,so that the release,dissemination and traceability of news content can be effectively managed.Then the study conducts deep learning processing of news through combining withBi-LSTM model,so as to improve the transparency of news and ensure its authenticity and credibility.