An Open-source Framework for Knowledge Embedding.
More information is available on our website http://openke.thunlp.org/
If you use the code, please cite the following paper:
@inproceedings{han2018openke,
title={OpenKE: An Open Toolkit for Knowledge Embedding},
author={Han, Xu and Cao, Shulin and Lv Xin and Lin, Yankai and Liu, Zhiyuan and Sun, Maosong and Li, Juanzi},
booktitle={Proceedings of EMNLP},
year={2018}
}
This is an Efficient implementation based on TensorFlow for knowledge representation learning (KRL). We use C++ to implement some underlying operations such as data preprocessing and negative sampling. For each specific model, it is implemented by TensorFlow with Python interfaces so that there is a convenient platform to run models on GPUs. OpenKE composes 4 repositories:
OpenKE: the main project based on TensorFlow, which provides the optimized and stable framework for knowledge graph embedding models.
OpenKE-PyTorch: OpenKE implemented with PyTorch, also providing the optimized and stable framework for knowledge graph embedding models.
TensorFlow-TransX: light and simple version of OpenKE based on TensorFlow, including TransE, TransH, TransR and TransD.
Fast-TransX: efficient lightweight C++ inferences for TransE and its extended models utilizing the framework of OpenKE, including TransH, TransR, TransD, TranSparse and PTransE.
Install TensorFlow
Clone the OpenKE repository:
$ git clone https://github.com/thunlp/OpenKE
$ cd OpenKE- Compile C++ files
$ bash make.shFor training, datasets contain three files:
train2id.txt: training file, the first line is the number of triples for training. Then the following lines are all in the format (e1, e2, rel) which indicates there is a relation rel between e1 and e2 . Note that train2id.txt contains ids from entitiy2id.txt and relation2id.txt instead of the names of the entities and relations. If you use your own datasets, please check the format of your training file. Files in the wrong format may cause segmentation fault.
entity2id.txt: all entities and corresponding ids, one per line. The first line is the number of entities.
relation2id.txt: all relations and corresponding ids, one per line. The first line is the number of relations.
For testing, datasets contain additional two files (totally five files):
test2id.txt: testing file, the first line is the number of triples for testing. Then the following lines are all in the format (e1, e2, rel) .
valid2id.txt: validating file, the first line is the number of triples for validating. Then the following lines are all in the format (e1, e2, rel) .
type_constrain.txt: type constraining file, the first line is the number of relations. Then the following lines are type constraints for each relation. For example, the relation with id 1200 has 4 types of head entities, which are 3123, 1034, 58 and 5733. The relation with id 1200 has 4 types of tail entities, which are 12123, 4388, 11087 and 11088. You can get this file through n-n.py in folder benchmarks/FB15K.
To compute a knowledge graph embedding, first import datasets and set configure parameters for training, then train models and export results. For instance, we write an example_train_transe.py to train TransE:
importconfigimportmodelsimporttensorflowastfimportnumpyasnpcon=config.Config()
#Input training files from benchmarks/FB15K/ folder.con.set_in_path("./benchmarks/FB15K/")
con.set_work_threads(4)
con.set_train_times(500)
con.set_nbatches(100)
con.set_alpha(0.001)
con.set_margin(1.0)
con.set_bern(0)
con.set_dimension(50)
con.set_ent_neg_rate(1)
con.set_rel_neg_rate(0)
con.set_opt_method("SGD")
#Models will be exported via tf.Saver() automatically.con.set_export_files("./res/model.vec.tf", 0)
#Model parameters will be exported to json files automatically.con.set_out_files("./res/embedding.vec.json")
#Initialize experimental settings.con.init()
#Set the knowledge embedding modelcon.set_model(models.TransE)
#Train the model.con.run()con.set_in_path("benchmarks/FB15K/")We import knowledge graphs from benchmarks/FB15K/ folder. The data consists of three essential files mentioned before:
- train2id.txt
- entity2id.txt
- relation2id.txt
Validation and test files are required and used to evaluate the training results, However, they are not indispensable for training.
con.set_work_threads(8)We can allocate several threads to sample positive and negative cases.
con.set_train_times(500)
con.set_nbatches(100)
con.set_alpha(0.5)
con.set_dimension(200)
con.set_margin(1)We set essential parameters, including the data traversing rounds, learning rate, batch size, and dimensions of entity and relation embeddings.
con.set_bern(0)
con.set_ent_neg_rate(1)
con.set_rel_neg_rate(0)For negative sampling, we can corrupt entities and relations to construct negative triples. set_bern(0) will use the traditional sampling method, and set_bern(1) will use the method in (Wang et al. 2014) denoted as "bern".
con.set_optimizer("SGD")We can select a proper gradient descent optimization algorithm to train models.
con.set_export_files("./res/model.vec.tf", 0)
con.set_out_files("./res/embedding.vec.json")Models will be exported via tf.Saver() automatically every few rounds. Also, model parameters will be exported to json files finally.
con.init()
con.set_model(models.TransE)
con.run()We set the knowledge graph embedding model and start the training process.
Link prediction aims to predict the missing h or t for a relation fact triple (h, r, t). In this task, for each position of missing entity, the system is asked to rank a set of candidate entities from the knowledge graph, instead of only giving one best result. For each test triple (h, r, t), we replace the head/tail entity by all entities in the knowledge graph, and rank these entities in descending order of similarity scores calculated by score function fr. we use two measures as our evaluation metric:
- MR : mean rank of correct entities;
- MRR: the average of the reciprocal ranks of correct entities;
- Hit@N : proportion of correct entities in top-N ranked entities.
Triple classification aims to judge whether a given triple (h, r, t) is correct or not. This is a binary classification task. For triple classification, we set a relationspecific threshold δr. For a triple (h, r, t), if the dissimilarity score obtained by fr is below δr, the triple will be classified as positive, otherwise negative. δr is optimized by maximizing classification accuracies on the validation set.
Given tail entity and relation, predict the top k possible head entities. All the objects are represented by their id.
defpredict_head_entity(self, t, r, k):
r'''This mothod predicts the top k head entities given tail entity and relation. Args: t (int): tail entity id r (int): relation id k (int): top k head entities Returns: list: k possible head entity ids '''self.init_link_prediction()
ifself.importName!=None:
self.restore_tensorflow()
test_h=np.array(range(self.entTotal))
test_r=np.array([r] *self.entTotal)
test_t=np.array([t] *self.entTotal)
res=self.test_step(test_h, test_t, test_r).reshape(-1).argsort()[:k]
print(res)
returnresThis is similar to predicting the head entity.
Given the head entity and tail entity, predict the top k possible relations. All the objects are represented by their id.
defpredict_relation(self, h, t, k):
r'''This methods predict the relation id given head entity and tail entity. Args: h (int): head entity id t (int): tail entity id k (int): top k relations Returns: list: k possible relation ids '''self.init_link_prediction()
ifself.importName!=None:
self.restore_tensorflow()
test_h=np.array([h] *self.relTotal)
test_r=np.array(range(self.relTotal))
test_t=np.array([t] *self.relTotal)
res=self.test_step(test_h, test_t, test_r).reshape(-1).argsort()[:k]
print(res)
returnresGiven a triple (h, r, t), this funtion tells us whether the triple is correct or not. If the threshold is not given, this function calculates the threshold for the relation from the valid dataset.
defpredict_triple(self, h, t, r, thresh=None):
r'''This method tells you whether the given triple (h, t, r) is correct of wrong Args: h (int): head entity id t (int): tail entity id r (int): relation id thresh (fload): threshold for the triple '''self.init_triple_classification()
ifself.importName!=None:
self.restore_tensorflow()
res=self.test_step(np.array([h]), np.array([t]), np.array([r]))
ifthresh!=None:
ifres<thresh:
print("triple (%d,%d,%d) is correct"% (h, t, r))
else:
print("triple (%d,%d,%d) is wrong"% (h, t, r))
returnself.lib.getValidBatch(self.valid_pos_h_addr, self.valid_pos_t_addr, self.valid_pos_r_addr, self.valid_neg_h_addr, self.valid_neg_t_addr, self.valid_neg_r_addr)
res_pos=self.test_step(self.valid_pos_h, self.valid_pos_t, self.valid_pos_r)
res_neg=self.test_step(self.valid_neg_h, self.valid_neg_t, self.valid_neg_r)
self.lib.getBestThreshold(self.relThresh_addr, res_pos.__array_interface__['data'][0], res_neg.__array_interface__['data'][0])
ifres<self.relThresh[r]:
print("triple (%d,%d,%d) is correct"% (h, t, r))
else: print("triple (%d,%d,%d) is wrong"% (h, t, r))To evaluate the model, first import datasets and set essential configure paramters, then set model parameters and test the model. For instance, we write an example_test_transe.py to test TransE.
There are four approaches to test models:
(1) Test models right after training.
importconfigimportmodelsimporttensorflowastfimportnumpyasnpcon=config.Config()
con.set_in_path("./benchmarks/FB15K/")
#True: Input test files from the same folder.con.set_test_triple_classification(True)
con.set_test_link_prediction(True)
con.set_work_threads(4)
con.set_train_times(500)
con.set_nbatches(100)
con.set_alpha(0.001)
con.set_margin(1.0)
con.set_bern(0)
con.set_dimension(50)
con.set_ent_neg_rate(1)
con.set_rel_neg_rate(0)
con.set_opt_method("SGD")
con.set_export_files("./res/model.vec.tf", 0)
con.set_out_files("./res/embedding.vec.json")
con.init()
con.set_model(models.TransE)
con.run()
#To test link prediction after training needs "set_test_link_prediction True)".#To test triple classfication after training needs "set_test_triple_classification(True)"con.test()(2) Set import files and OpenKE will automatically load models via tf.Saver().
importconfigimportmodelsimporttensorflowastfimportnumpyasnpimportjsoncon=config.Config()
con.set_in_path("./benchmarks/FB15K/")
con.set_test_link_prediction(True)
con.set_test_triple_classification(True)
con.set_work_threads(4)
con.set_dimension(50)
con.set_import_files("./res/model.vec.tf")
con.init()
con.set_model(models.TransE)
con.test()(3) Read model parameters from json files and manually load parameters.
importconfigimportmodelsimporttensorflowastfimportnumpyasnpimportjsoncon=config.Config()
con.set_in_path("./benchmarks/FB15K/")
con.set_test_link_prediction(True)
con.set_test_triple_classification(True)
con.set_work_threads(4)
con.set_dimension(50)
con.init()
con.set_model(models.TransE)
f=open("./res/embedding.vec.json", "r")
content=json.loads(f.read())
f.close()
con.set_parameters(content)
con.test()(4) Manually load models via tf.Saver().
importconfigimportmodelsimporttensorflowastfimportnumpyasnpimportjsoncon=config.Config()
con.set_in_path("./benchmarks/FB15K/")
con.set_test_link_prediction(True)
con.set_test_triple_classification(True)
con.set_work_threads(4)
con.set_dimension(50)
con.init()
con.set_model(models.TransE)
con.import_variables("./res/model.vec.tf")
con.test()Note that you can only load model parameters when model configuration finished.
There are four approaches to get the embedding matrix.
(1) Set import files and OpenKE will automatically load models via tf.Saver().
con=config.Config()
con.set_in_path("./benchmarks/FB15K/")
con.set_test_link_prediction(True)
con.set_test_triple_classification(True)
con.set_work_threads(4)
con.set_dimension(50)
con.set_import_files("./res/model.vec.tf")
con.init()
con.set_model(models.TransE)
# Get the embeddings (numpy.array)embeddings=con.get_parameters("numpy")
# Get the embeddings (python list)embeddings=con.get_parameters()(2) Read model parameters from json files and manually load parameters.
con=config.Config()
con.set_in_path("./benchmarks/FB15K/")
con.set_test_link_prediction(True)
con.set_test_triple_classification(True)
con.set_work_threads(4)
con.set_dimension(50)
con.init()
con.set_model(models.TransE)
f=open("./res/embedding.vec.json", "r")
embeddings=json.loads(f.read())
f.close()(3) Manually load models via tf.Saver().
con=config.Config()
con.set_in_path("./benchmarks/FB15K/")
con.set_test_link_prediction(True)
con.set_test_triple_classification(True)
con.set_work_threads(4)
con.set_dimension(50)
con.init()
con.set_model(models.TransE)
con.import_variables("./res/model.vec.tf")
# Get the embeddings (numpy.array)embeddings=con.get_parameters("numpy")
# Get the embeddings (python list)embeddings=con.get_parameters()(4) Immediately get the embeddings after training the model.
...
...
...
#Models will be exported via tf.Saver() automatically.con.set_export_files("./res/model.vec.tf", 0)
#Model parameters will be exported to json files automatically.con.set_out_files("./res/embedding.vec.json")
#Initialize experimental settings.con.init()
#Set the knowledge embedding modelcon.set_model(models.TransE)
#Train the model.con.run()
#Get the embeddings (numpy.array)embeddings=con.get_parameters("numpy")
#Get the embeddings (python list)embeddings=con.get_parameters()classConfig(object):
#To set the learning ratedefset_alpha(alpha=0.001)
#To set the degree of the regularization on the parametersdefset_lmbda(lmbda=0.0)
#To set the gradient descent optimization algorithm (SGD, Adagrad, Adadelta, Adam)defset_optimizer(optimizer="SGD")
#To set the data traversing roundsdefset_train_times(self, times)
#To split the training triples into several batches, nbatches is the number of batchesdefset_nbatches(nbatches=100)
#To set the margin for the loss functiondefset_margin(margin=1.0)
#To set the dimensions of the entities and relations at the same timedefset_dimension(dim)
#To set the dimensions of the entitiesdefset_ent_dimension(self, dim)
#To set the dimensions of the relationsdefset_rel_dimension(self, dim)
#To allocate threads for each batch samplingdefset_work_threads(threads=1)
#To set negative sampling algorithms, unif (bern = 0) or bern (bern = 1)defset_bern(bern=1)
#For each positive triple, we construct rate negative triples by corrupt the entitydefset_ent_neg_rate(rate=1)
#For each positive triple, we construct rate negative triples by corrupt the relationdefset_rel_neg_rate(rate=0)
#To sample a batch of training triples, including positive and negative ones.defsampling()
#To import dataset from the benchmark folderdefset_in_path(self, path)
#To export model parameters to json files when training completeddefset_out_files(self, path)
#To set the import files, all parameters can be restored from the import filesdefset_import_files(self, path)
#To set the export file of model paramters, and export results every few roundsdefset_export_files(self, path, steps=0)
#To export results every few roundsdefset_export_steps(self, steps)
#To save model via tf.saverdefsave_tensorflow(self)
#To restore model via tf.saverdefrestore_tensorflow(self)
#To export model paramters, when path is none, equivalent to save_tensorflow()defexport_variables(self, path=None)
#To import model paramters, when path is none, equivalent to restore_tensorflow()defimport_variables(self, path=None)
#To export model paramters to designated pathdefsave_parameters(self, path=None)
#To manually load parameters which are read from json filesdefset_parameters(self, lists)
#To get model paramters, if using mode "numpy", you can get np.array , else you can get python listsdefget_parameters(self, mode="numpy")
#To set the knowledge embedding modeldefset_model(model)
#The framework will print loss values during training if flag = 1defset_log_on(flag=1)
#This is essential when testingdefset_test_link_prediction(True)
defset_test_triple_classification(True)classModel(object)
# return config which saves the training parameters.get_config(self)
# in_batch = True, return [positive_head, positive_tail, positive_relation]# The shape of positive_head is [batch_size, 1]# in_batch = False, return [positive_head, positive_tail, positive_relation]# The shape of positive_head is [batch_size]get_positive_instance(in_batch=True)
# in_batch = True, return [negative_head, negative_tail, negative_relation]# The shape of positive_head is [batch_size, negative_ent_rate + negative_rel_rate]# in_batch = False, return [negative_head, negative_tail, negative_relation]# The shape of positive_head is [(negative_ent_rate + negative_rel_rate) * batch_size] get_negative_instance(in_batch=True)
# in_batch = True, return all training instances with the shape [batch_size, (1 + negative_ent_rate + negative_rel_rate)]# in_batch = False, return all training instances with the shape [(negative_ent_rate + negative_rel_rate + 1) * batch_size]defget_all_instance(in_batch=False)
# in_batch = True, return all training labels with the shape [batch_size, (1 + negative_ent_rate + negative_rel_rate)]# in_batch = False, return all training labels with the shape [(negative_ent_rate + negative_rel_rate + 1) * batch_size]# The positive triples are labeled as 1, and the negative triples are labeled as -1defget_all_labels(in_batch=False)
# To define containers for training triplesdefinput_def()
# To define embedding parameters for knowledge embedding modelsdefembedding_def()
# To define loss functions for knowledge embedding modelsdefloss_def()
# To define the prediction functions for knowledge embedding modelsdefpredict_def(self)
def__init__(config)
#The implementation for TransEclassTransE(Model)
#The implementation for TransH classTransH(Model)
#The implementation for TransRclassTransR(Model)
#The implementation for TransDclassTransD(Model)
#The implementation for RESCALclassRESCAL(Model)
#The implementation for DistMultclassDistMult(Model)
#The implementation for HolEclassHolE(Model) #The implementation for ComplExclassComplEx(Model)