{
  "fetched_at": "2026-10-19T14:05:14Z",
  "source_url": "https://www.kaggle.com/competitions/cs245-2-project-2-key-author-finding/data",
  "pages": [
    {
      "name": "data-description",
      "content": "## File descriptions\n\n*   **author_info_covid.json** - Contains a dictionary, for each key is an anonymous author id, and each value is the basic information in the format of `dict` for the author. For each author, we provide the nationality of the author and the fields the author has published a paper in. The detailed format and sample are as follows.\n\n\n> format\n\n> 'author_id' : {\n\n>     \"ct\": \"country code with 2 letters\",\n\n>     \"ca\": [[\"anonymous field id\", count_for_the_papers]]\n\n> }\n\n> sample\n\n> '0' : {\n\n>     \"ct\": \"SE\",\n\n>     \"ca\": [\n\n>         [\"11\", 18],\n\n>         [\"06\", 1]]\n\n> }\n\n> \"\"\"\n\n> Author with id \"0\" comes from the country name \"SE\" and has published \"18\" papers in the field \"11\" and only \"1\" paper in the field \"1\".\n\n> \"\"\"\n\n\n*   **coauthor_covid.txt** - Contains an anonymous coauthor network with three integers per line, denoted as (a1, a2, t), which means that author a1 and author a2 has cooperated for t times.\n\n*   **authors_to_pred.csv** - The list of author id that need to be predicted, and it's recommended to output your prediction in the order of id in this file. (Test set)\n",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Description",
      "content": "<p>PLEASE JOIN THIS COMPETITION BEFORE 2021-05-15</p>\n<h2>Problem Description</h2>\n\nDuring the pandemic, many scientific researchers have devoted themselves to the pathology, transmission, policy, control methods, and other aspects of the epidemic. In addition to biomedical researchers, there are also many scientific researchers from other disciplines who use their related work to analyze and model. and solve the problems caused by the pandemic. It has spawned many interdisciplinary collaborations, cross-domain projects, and academic achievements. \n\nIn scholarly network analysis, there are many issues worthy of our attention. Structurally, the network model is divided into the regular network, random network, small-world network, and a scale-free network. We generally think of it as a scale-free network in the scholar network. In the face of the cooperative relationship between scholars, we often use the clustering coefficient to measure it. By calculating the degree of interconnection between the neighbor nodes of a point, we can know the degree of mutual understanding between scholars. We can use this indicator during the epidemic to know how close the cooperation between scholars is. Related to this is the sparsity of the network, and the cooperation networks in different fields have different sparsity.\n\nNext, we can further analyze the network's topology through different motifs. Some cooperation models are radial, which means that a strong scholar leads some lesser scholars, and some cooperation models are chained, representing scholars. They do not know each other, but they get what they need. There is also a cooperation model that is two radially connected, relying on two influential scholars to cooperate to combine the two big teams. These scholars who have met and cooperated with two initially unknown scholars are identified as critical scholars, and these scholars play an essential role in scientific research.\n\nWe can sometimes think that the scholar cooperation network is a small-world network. Such a network often has a smaller average distance and a more significant clustering coefficient. These indicators are all used to measure a community, but How much influence does a point have on the clustering and connectivity of this community? Currently, only centrality indicators can be measured,\n\nWe can use betweenness centrality (in the shortest paths of all pairs of nodes in the network, the shortest paths passing through a node, the more influential the node is), closeness centrality (the average distance a node has from other nodes in the network, the more is smaller, the greater the closeness of the node is), the eigenvector centerline (the importance of a node depends on both the number of its neighbors (that is, the degree of the node) and the importance of each neighbor), degree centrality (the more neighbors a node has, the more important it is), semi-local centrality (count the number of second-order neighbors). However, this indicator ignores the attributes of scholars themselves. Therefore, it is far from enough to identify a critical scholar. Therefore, we take the opportunity of the big assignment to discuss the classification of this critical scholar with you.\n\nThe data we give contains:\n\n1. A scholar's institution, country, and discipline, from which we can know an essential attribute of a scholar. \n2. When two scholars come from the same institution and are in the same discipline, the cooperation situation is likely to be intra-group cooperation.\n3. When two scholars come from different institutions but are still in the same discipline, the cooperation situation is likely to be the cooperation of disciplines.\n4. If two scholars come from different countries and disciplines, then the two scholars are engaged in some interdisciplinary, international cooperation, which is likely to be a specific topic that requires joint exploration by multiple countries.\n\nWe divide these scholars into four categories:\n\n1. the heart of the network\n2. network switch\n3. network router\n4. network edge node\n\nplease design an indicator that can divide the scholars in the network.\n\n<h2>Attention</h2>\n\nBefore submitting your predictions, please update your name with \"GroupNumber_StudentName\"  before submitting your predictions. You can update your name in the \"Team\" tab.\n\nFor fairness, we set a few rules for you to obey --- violation against the rules will lead to score \ndeduction:\n1. Please do not copy someone else's code. We will run plagiarism check after \nsubmission.\n2. Please do not download the dataset somewhere else to train your model. We have \nshuffled the dataset and will reproduce your experimental results using the dataset. A \nviolation would be considered if there is a big gap between your reported results and our \nreproduced results.\n3. Please do not use the pre-trained model. Everything should be built from scratch.\n\n<h2>Acknowledgements</h2>\n<p>Data used in this project is provided by <a href=\"https://www.acemap.info/\">Acemap </a>group.</p>",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Evaluation",
      "content": "<hr>\n<p>The evaluation metric for this competition is <a href=\"https://en.wikipedia.org/wiki/Pearson_correlation_coefficient\"> Pearson Correlation Coefficient</a>(PCC). PCC is a measure of linear correlation between two sets of data. It is the ratio between the covariance of two variables and the product of their standard deviations, and is given by:</p>\n<p>PCC(X,Y) = \\frac{cov(X,Y)}{\\sigma_X \\sigma_Y}</p>\n<p>where cov(X,Y) denotes the covariance of X and Y, \\sigma_X, \\sigma_Y is the standard deviation of X and Y, respectively.</p>\n<p>\nBefore the submission deadline , you can only view the PCC evaluated on 50% of the dataset on the leaderboard. After the deadline, we will reveal your model PCC evaluated on the entire dataset to rate your performance.\n</p>\n<h2>Submission Format</h2>\n<p>Your submission file should contain two columns: id and pred_label. Each test author node  is allocated a unique id, and you need to use this id as your prediction index. For example, the first element in the first row <strong>0</strong> means a test node with author id 0. If you predict that it is a key author in the network, the value of the corresponding row of your uploaded file is 0,1. \n</p><p>The file should contain a header and have the following format:</p>\n<pre>id,pred_label<br>0,1<br>1,0<br>17,0<br></pre>\n<p>Detailed description can be found in ‘data’ page.</p>\n",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "rules",
      "content": "1. Please do not copy someone else's code. We will run plagiarism check after submission.\n\n2. Please do not download the dataset somewhere else to train your model. We have shuffled the dataset and will reproduce your experimental results using the dataset. A violation would be considered if there is a big gap between your reported results and our reproduced results.\n\n3. Please do not use the pre-trained model. Everything should be built from scratch.\n\n4. You need to zip and upload the following files before the deadline: (1) Your codes. (2) A readme.txt, which includes the scripts or instructions to run your codes. (3) A research paper, write in conference format. The final zip file should be uploaded via our course website: [https://oc.sjtu.edu.cn](https://oc.sjtu.edu.cn)",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    },
    {
      "name": "Reference",
      "content": "1. 【PNAS】Consolidation in a crisis: Patterns of international collaboration in early COVID-19 research. [paper](https://journals.plos.org/plosone/article/file?id=10.1371/journal.pone.0236307&type=printable)\n2. 【PNAS】Collaboration and Knowledge Networks: A Framework on Analyzing Evolution of University-industry Collaborative Innovation. [paper](http://ceur-ws.org/Vol-2871/paper3.pdf)\n3. 【Nature Communication】Dynamics of social network emergence explain network evolution.  [paper](https://www.nature.com/articles/s41598-020-78224-2.pdf)\n4. 【Financial Science】The Evolving Network of Legal Scholars. [paper](https://papers.ssrn.com/sol3/Delivery.cfm/SSRN_ID3990267_code610406.pdf?abstractid=3833993&mirid=1)\n5. 【Scientific Report】Ollivier-Ricci Curvature-Based Method to Community Detection in Complex Networks. [paper](https://www.nature.com/articles/s41598-019-46079-x.pdf)",
      "isPublished": false,
      "hasIsPublished": false,
      "postTitle": "",
      "hasPostTitle": false,
      "mimeType": "",
      "hasMimeType": false
    }
  ]
}