Combination of feature engineering and ranking models for paper-author identification in KDD Cup 2013

Li, Chun-Liang; Su, Yu-Chuan; Lin, Ting-Wei; Tsai, Chia-Liang; Chang, Wei‐Cheng; Huang, Kuan-Hao; Kuo, Tsun-Cheng; Lin, Shan-Wei; Lin, Young-San; Lu, Yuchen; Yang, Chun‐Pai; Chang, Chun-Ti; Chin, Wei-Sheng; Juan, Yu-Chin; Tung, Hsiao-Yu Fish; Wang, Jui-pin; Wei, Cheng-Kuang; Wu, Felix; Yin, Tu-Chun; Yu, Tong; Zhuang, Yun; Lin, Shou-De; Lin, Hung‐Mo

doi:10.1145/2517288.2517290

Cited by 15 publications

(10 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…• Sup: Triggered by KDD Cup 2013, the problem of author identification has recently garnered attention, and top solutions of the challenge heavily relied on feature engineering followed by supervised ranking models on these features [8,17]. Following them, we extract 16 features for each pair of paper and author in the training set.…”

Section: Methodsmentioning

confidence: 99%

Task-Guided Pair Embedding in Heterogeneous Network

Park

Kim

Zhu

et al. 2019

Proceedings of the 28th ACM International Conference on Information and Knowledge Management

View full text Add to dashboard Cite

Many real-world tasks solved by heterogeneous network embedding methods can be cast as modeling the likelihood of a pairwise relationship between two nodes. For example, the goal of author identification task is to model the likelihood of a paper being written by an author (paper-author pairwise relationship). Existing taskguided embedding methods are node-centric in that they simply measure the similarity between the node embeddings to compute the likelihood of a pairwise relationship between two nodes. However, we claim that for task-guided embeddings, it is crucial to focus on directly modeling the pairwise relationship. In this paper, we propose a novel task-guided pair embedding framework in heterogeneous network, called TaPEm, that directly models the relationship between a pair of nodes that are related to a specific task (e.g., paper-author relationship in author identification). To this end, we 1) propose to learn a pair embedding under the guidance of its associated context path, i.e., a sequence of nodes between the pair, and 2) devise the pair validity classifier to distinguish whether the pair is valid with respect to the specific task at hand. By introducing pair embeddings that capture the semantics behind the pairwise relationships, we are able to learn the fine-grained pairwise relationship between two nodes, which is paramount for task-guided embedding methods. Extensive experiments on author identification task demonstrate that TaPEm outperforms the state-of-the-art methods, especially for authors with few publication records.

show abstract

Section: Methodsmentioning

confidence: 99%

Task-Guided Pair Embedding in Heterogeneous Network

Park

Kim

Zhu

et al. 2019

Proceedings of the 28th ACM International Conference on Information and Knowledge Management

View full text Add to dashboard Cite

show abstract

“…The problem of author identification has been briefly studied before [11]. And we also notice KDD Cup 2013 has similar author identification/disambiguation problem [12,13,34,9,33], where participants are asked to predict which paper is truly written by some author. However, different from the KDD Cup, our setting is different from them in the sense that (1) existing authors are unknown in our double-blind setting, and (2) we consider the reference of the paper, which is one of the most important sources of information.…”

Section: Related Workmentioning

confidence: 99%

“…Unlike traditional supervised learning, dense vectorized representations [16,15] are not directly available in networked data [26]. Hence, many traditional methods under network settings heavily rely on problem specific feature engineering [12,13,34,9,33].Although feature engineering can incorporate prior knowledge of the problem and network structure, usually it is time-consuming, problem specific (thus not transferable), and the extracted features may be too simple for complicated data sets [3]. Several network embedding methods [17,26,25] have been proposed to automatically learn feature representations for networked data.…”

mentioning

confidence: 99%

Task-Guided and Path-Augmented Heterogeneous Network Embedding for Author Identification

Chen

Sun

2017

Proceedings of the Tenth ACM International Conference on Web Search and Data Mining

208

165

View full text Add to dashboard Cite

In this paper, we study the problem of author identification under double-blind review setting, which is to identify potential authors given information of an anonymized paper. Different from existing approaches that rely heavily on feature engineering, we propose to use network embedding approach to address the problem, which can automatically represent nodes into lower dimensional feature vectors. However, there are two major limitations in recent studies on network embedding: (1) they are usually general-purpose embedding methods, which are independent of the specific tasks; and (2) most of these approaches can only deal with homogeneous networks, where the heterogeneity of the network is ignored. Hence, challenges faced here are two folds: (1) how to embed the network under the guidance of the author identification task, and (2) how to select the best type of information due to the heterogeneity of the network.To address the challenges, we propose a task-guided and pathaugmented heterogeneous network embedding model. In our model, nodes are first embedded as vectors in latent feature space. Embeddings are then shared and jointly trained according to task-specific and network-general objectives. We extend the existing unsupervised network embedding to incorporate meta paths in heterogeneous networks, and select paths according to the specific task. The guidance from author identification task for network embedding is provided both explicitly in joint training and implicitly during meta path selection. Our experiments demonstrate that by using pathaugmented network embedding with task guidance, our model can obtain significantly better accuracy at identifying the true authors comparing to existing methods. network mining problems, good representations of data are very important, as demonstrated by many previous work [16,15,17,26,7]. Unlike traditional supervised learning, dense vectorized representations [16,15] are not directly available in networked data [26]. Hence, many traditional methods under network settings heavily rely on problem specific feature engineering [12,13,34,9,33].Although feature engineering can incorporate prior knowledge of the problem and network structure, usually it is time-consuming, problem specific (thus not transferable), and the extracted features may be too simple for complicated data sets [3]. Several network embedding methods [17,26,25] have been proposed to automatically learn feature representations for networked data. A key idea behind network embedding is learning to map nodes into vector space, such that the proximities among nodes can be preserved. Similar nodes (in terms of connectivity, or other properties) are expected to be placed near each other in the vector space.Unfortunately, most existing embedding methods produce generalpurpose embeddings that are independent of tasks, and they are usually designed for homogeneous networks [17,26]. When it comes to author identification problem under the heterogeneous networks, existing embedding methods cannot be applied direc...

show abstract

“…In organizational networks, graph-based models, largely based on random walk [23], are widely used to estimate individual authority. In this field, extensive researches have demonstrated strong correlations between centrality and authority [10,16,26,27]. The famous PageRank algorithm proposed by L. Page et al [18] and later the Topic-sensitive Pagerank [11] have proven the value of the citation graph for web pages.…”

Section: Related Workmentioning

confidence: 99%

A Unified Approach for Learning Expertise and Authority in Digital Libraries

Robertie

Ermakova

Pitarch

et al. 2017

Database Systems for Advanced Applications

View full text Add to dashboard Cite

OATAO is an open access repository that collects the work of Toulouse researchers and makes it freely available over the web where possible. This is an author-deposited version published in : http://oatao.univ-toulouse.fr/ Eprints ID : 18909The contribution was presented at DASFAA 2017 :http://ada.suda.edu.cn/dasfaa2017/ Abstract. Managing individual expertise is a major concern within any industrial-wide organization. If previous works have extensively studied the related expertise and authority profiling issues, they assume a semantic independence of these two key concepts. In digital libraries, stateof-the-art models generally summarize the researchers' profile by using solely textual information. Consequently, authors with a large amount of publications are mechanically fostered to the detriment of less prolific ones with probably higher expertise. To overcome this drawback we propose to merge the two representations of expertise and authority and balance the results by capturing a mutual reinforcement principle between these two notions. Based on a graph representation of the library, the expert profiling task is formulated as an optimization problem where latent expertise and authority representations are learned simultaneously, unbiasing the expertise scores of individuals with a large amount of publications. The proposal is instanciated on a public scientific bibliographic dataset where researchers' publications are considered as a source of evidence of individuals' expertise and citation relations as a source of authoritative signals. Results from our experiments conducted over the Microsoft Academic Search database demonstrate significant efficiency improvement in comparison with state-of-the-art models for the expert retrieval task.

show abstract

Combination of feature engineering and ranking models for paper-author identification in KDD Cup 2013

Cited by 15 publications

References 12 publications

Task-Guided Pair Embedding in Heterogeneous Network

Task-Guided Pair Embedding in Heterogeneous Network

Task-Guided and Path-Augmented Heterogeneous Network Embedding for Author Identification

A Unified Approach for Learning Expertise and Authority in Digital Libraries

Contact Info

Product

Resources

About