{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/task-guided-and-path-augmented-heterogeneous","title":"Task-Guided and Path-Augmented Heterogeneous Network Embedding for Author Identification","arxiv_id":"1612.02814","date":"2016-12-08","proceeding":null,"authors":["Ting Chen","Yizhou Sun"],"abstract":"In this paper, we study the problem of author identification under\ndouble-blind review setting, which is to identify potential authors given\ninformation of an anonymized paper. Different from existing approaches that\nrely heavily on feature engineering, we propose to use network embedding\napproach to address the problem, which can automatically represent nodes into\nlower dimensional feature vectors. However, there are two major limitations in\nrecent studies on network embedding: (1) they are usually general-purpose\nembedding methods, which are independent of the specific tasks; and (2) most of\nthese approaches can only deal with homogeneous networks, where the\nheterogeneity of the network is ignored. Hence, challenges faced here are two\nfolds: (1) how to embed the network under the guidance of the author\nidentification task, and (2) how to select the best type of information due to\nthe heterogeneity of the network.\n  To address the challenges, we propose a task-guided and path-augmented\nheterogeneous network embedding model. In our model, nodes are first embedded\nas vectors in latent feature space. Embeddings are then shared and jointly\ntrained according to task-specific and network-general objectives. We extend\nthe existing unsupervised network embedding to incorporate meta paths in\nheterogeneous networks, and select paths according to the specific task. The\nguidance from author identification task for network embedding is provided both\nexplicitly in joint training and implicitly during meta path selection. Our\nexperiments demonstrate that by using path-augmented network embedding with\ntask guidance, our model can obtain significantly better accuracy at\nidentifying the true authors comparing to existing methods.","url_abs":"http://arxiv.org/abs/1612.02814v2","url_pdf":"http://arxiv.org/pdf/1612.02814v2.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"task-guided-and-path-augmented-heterogeneous","repo_url":"https://github.com/chentingpc/GuidedHeteEmbedding","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":null}],"tasks":[{"task_slug":"feature-engineering","task_name":"Feature Engineering"},{"task_slug":"network-embedding","task_name":"Network Embedding"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}