DRMM: A Deep Relevance Matching Model for Ad-hoc Retrieval.MatchPyramid: Text Matching as Image RecognitionARC-I: Convolutional Neural Network Architectures for Matching Natural Language SentencesDSSM: Learning Deep Structured Semantic Models for Web Search using Clickthrough DataCDSSM: Learning Semantic Representations Using Convolutional Neural Networks for Web SearchARC-II: Convolutional Neural Network Architectures for Matching Natural Language SentencesMV-LSTM: A Deep Architecture for Semantic Matching with Multiple Positional Sentence RepresentationsaNMM: aNMM: Ranking Short Answer Texts with Attention-Based Neural Matching ModelDUET: Learning to Match Using Local and Distributed Representations of Text for Web SearchK-NRM: End-to-End Neural Ad-hoc Ranking with Kernel PoolingCONV-KNRM: Convolutional neural networks for soft-matching n-grams in ad-hoc search
基本的解決框架是：
第一步：詞表示\上下文表示
第二步： 詞條表示
第三步：相似度計算
第四步：選出相關性得分最高的詞條
沒有訓練資料的話，直接無監督，把上下文和詞條分詞後詞向量加起來，複雜點用BERT編碼。然後直接用計算餘弦相似度。
有訓練資料的話，可以把正確的詞條當正例，其他詞條當負例，然後用文字匹配模型訓練一波。深度學習的文字匹配模型有很多，提供一些給你做參考：
DRMM: A Deep Relevance Matching Model for Ad-hoc Retrieval.MatchPyramid: Text Matching as Image RecognitionARC-I: Convolutional Neural Network Architectures for Matching Natural Language SentencesDSSM: Learning Deep Structured Semantic Models for Web Search using Clickthrough DataCDSSM: Learning Semantic Representations Using Convolutional Neural Networks for Web SearchARC-II: Convolutional Neural Network Architectures for Matching Natural Language SentencesMV-LSTM: A Deep Architecture for Semantic Matching with Multiple Positional Sentence RepresentationsaNMM: aNMM: Ranking Short Answer Texts with Attention-Based Neural Matching ModelDUET: Learning to Match Using Local and Distributed Representations of Text for Web SearchK-NRM: End-to-End Neural Ad-hoc Ranking with Kernel PoolingCONV-KNRM: Convolutional neural networks for soft-matching n-grams in ad-hoc search