graph TD;
    A[Building Sememe Knowlage Base via Unsupervised Learning]-->B[Sememe Knowlage Base for Cross-lingual Language Model];
    B[Sememe Knowlage Base for Cross-lingual Language Model]-->C[Cross-lingual Textual Information Isomorphism for Conversational Recommendation System];
    D[Paradigm unification for Pre-training]-->E[Textual Information Isomorphism for Pre-training];
    E[Textual Information Isomorphism for Pre-training]-->C[Cross-lingual Textual Information Isomorphism for Conversational Recommendation System];
    F[Dialogue Systems using Reinforcement Learning]-->G[Conversational Recommendation System];
    G[Conversational Recommendation System]-->C[Cross-lingual Textual Information Isomorphism for Conversational Recommendation System];

• Cross-lingual Textual Information Isomorphism for Conversational Recommendation System
  • Sememe Knowlage Base for Cross-lingual Language Model
    • Building Sememe Knowlage Base via Unsupervised Learning
  • Textual Information Isomorphism for Pre-training
    • Paradigm unification for Pre-training
  • Conversational Recommendation System
    • Dialogue Systems using Reinforcement Learning

---