TransNet: Translation-Based Network Representation Learning for Social Relation Extraction
Authors: Cunchao Tu, Zhengyan Zhang, Zhiyuan Liu, Maosong Sun
IJCAI 2017 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Experimental results on SRE demonstrate that Trans Net significantly outperforms other baseline methods by 10% to 20% on hits@1. The source code and datasets can be obtained from https: //github.com/thunlp/Trans Net. |
| Researcher Affiliation | Academia | Cunchao Tu1,2, Zhengyan Zhang1, Zhiyuan Liu1,2 , Maosong Sun1,2 1Department of Computer Science and Technology, State Key Lab on Intelligent Technology and Systems, National Lab for Information Science and Technology, Tsinghua University, China 2Jiangsu Collaborative Innovation Center for Language Ability, Jiangsu Normal University, China {tcc13, zhangzhengyan14}@mails.tsinghua.edu.cn, liuzy@tsinghua.edu.cn, sms@mail.tsinghua.edu.cn |
| Pseudocode | No | No pseudocode or algorithm blocks were found in the paper. |
| Open Source Code | Yes | The source code and datasets can be obtained from https: //github.com/thunlp/Trans Net. |
| Open Datasets | Yes | The source code and datasets can be obtained from https: //github.com/thunlp/Trans Net. Firstly, we collect all the research interest phrases from the author profiles and build the label vocabulary with these phrases. These phrases are mainly crawled from the authors personal home pages and annotated by themselves. Hence, these phrases are rather credible, which is also confirmed by our manual check. Secondly, for each co-author relationship, we filter out the in-vocabulary labels in the abstracts of coauthored papers and regard them as the ground truth labels of this edge. Note that, as the edges in co-author networks are undirected, we replace each edge with two directed edges with opposite directions. Specifically, to better investigate the characteristics of different models, we construct three datasets with different scales, denoted as Arnet-S(small), Arnet-M(medium) and Arnet-L(large). The details are shown in Table 1. Table 1: Datasets. (ML indicates multi-label edges.) Datasets Arnet-S Arnet-M Arnet-L Vertices 187, 939 268, 037 945, 589 Edges 1, 619, 278 2, 747, 386 5, 056, 050 Train 1, 579, 278 2, 147, 386 3, 856, 050 Test 20, 000 300, 000 600, 000 Valid 20, 000 300, 000 600, 000 Labels 100 500 500 ML Proportion (%) 42.46 63.74 61.68 |
| Dataset Splits | Yes | Table 1: Datasets. (ML indicates multi-label edges.) Datasets Arnet-S Arnet-M Arnet-L Vertices 187, 939 268, 037 945, 589 Edges 1, 619, 278 2, 747, 386 5, 056, 050 Train 1, 579, 278 2, 147, 386 3, 856, 050 Test 20, 000 300, 000 600, 000 Valid 20, 000 300, 000 600, 000 Labels 100 500 500 ML Proportion (%) 42.46 63.74 61.68 |
| Hardware Specification | No | No specific hardware details (e.g., CPU, GPU models, memory, cloud instances) used for experiments were mentioned in the paper. |
| Software Dependencies | No | The paper mentions using the "Adam algorithm" and "dropout" but does not provide specific version numbers for any software libraries or dependencies (e.g., Python, TensorFlow, PyTorch, scikit-learn versions). |
| Experiment Setup | Yes | We set the representation dimension to 100 for all models. In Trans Net, we set the regularizer weight η to 0.001, the learning rate to 0.001 and the margin γ to 1. Besides, we employ a 2-layer autoencoder for all datasets and select bestperformed hyper-parameters α and β on validation sets. At last, we adopt Adam algorithm [Kingma and Ba, 2015] to minimize the objective in Eq. (7). In order to prevent overfitting, we also employ dropout [Srivastava et al., 2014] to generate the edge representations. |