MatchZoo 文本匹配工具包

MatchZoo是封装了一系列文本匹配的框架包含以下算法：

官网地址：https://github.com/kouunn/MatchZoo

Model Detail:

1. DRMM

this model is an implementation of A Deep Relevance Matching Model for Ad-hoc Retrieval.

• model file: models/drmm.py
• model config: models/drmm_ranking.config

---

2. MatchPyramid

this model is an implementation of Text Matching as Image Recognition

• model file: models/matchpyramid.py
• model config: models/matchpyramid_ranking.config

---

3. ARC-I

this model is an implementation of Convolutional Neural Network Architectures for Matching Natural Language Sentences

• model file: models/arci.py
• model config: models/arci_ranking.config

---

4. DSSM

this model is an implementation of Learning Deep Structured Semantic Models for Web Search using Clickthrough Data

• model file: models/dssm.py
• model config: models/dssm_ranking.config

---

5. CDSSM

this model is an implementation of Learning Semantic Representations Using Convolutional Neural Networks for Web Search

• model file: models/cdssm.py
• model config: models/cdssm_ranking.config

---

6. ARC-II

this model is an implementation of Convolutional Neural Network Architectures for Matching Natural Language Sentences

• model file: models/arcii.py
• model config: models/arcii_ranking.config

---

7. MV-LSTM

this model is an implementation of A Deep Architecture for Semantic Matching with Multiple Positional Sentence Representations

• model file: models/mvlstm.py
• model config: models/mvlstm_ranking.config

---

8. aNMM

this model is an implementation of aNMM: Ranking Short Answer Texts with Attention-Based Neural Matching Model

• model file: models/anmm.py
• model config: models/anmm_ranking.config

---

9. DUET

this model is an implementation of Learning to Match Using Local and Distributed Representations of Text for Web Search

• model file: models/duet.py
• model config: models/duet_ranking.config

---

10. K-NRM

this model is an implementation of End-to-End Neural Ad-hoc Ranking with Kernel Pooling

• model file: models/knrm.py
• model config: models/knrm_ranking.config

---

11. CONV-KNRM:

this model is an implementation of Convolutional neural networks for soft-matching n-grams in ad-hoc search

• model file: models/convknrm.py
• model config: models/convknrm.config

一些不同模型的测试成绩:

Models	NDCG@3	NDCG@5	MAP
DSSM	0.5439	0.6134	0.5647
CDSSM	0.5489	0.6084	0.5593
ARC-I	0.5680	0.6317	0.5870
ARC-II	0.5647	0.6176	0.5845
MV-LSTM	0.5818	0.6452	0.5988
DRMM	0.6107	0.6621	0.6195
K-NRM	0.6268	0.6693	0.6256
aNMM	0.6160	0.6696	0.6297
DUET	0.6065	0.6722	0.6301
MatchPyramid	0.6317	0.6913	0.6434
DRMM_TKS	0.6458	0.6956	0.6586

 

Environment

• python2.7+
• tensorflow 1.2+
• keras 2.06+
• nltk 3.2.2+
• tqdm 4.19.4+
• h5py 2.7.1+