停用高级电子数据展示中的相关性模块Retirement of the Relevance module in Advanced eDiscovery

2021 年 3 月 10 日,我们将停用高级电子数据展示中的相关性模块。On March 10, 2021, we are retiring the Relevance module in Advanced eDiscovery. 此停用意味着组织将无法再访问相关性模块 (,方法为在高级电子数据展示案例) 中管理审阅集相关性,或能够访问任何现有相关性 > 模型。This retirement means that organizations will no longer have access to the Relevance module (by going to Manage review set > Relevance in an Advanced eDiscovery case) or be able to access any existing Relevance models. 即将停用的当前相关性模块将在 2021 年第 2 季度替换为新的预测编码解决方案。The current Relevance module that is being retired will be replaced with a new predictive coding solution in Q2 CY 2021. 这一新功能将允许组织在更轻松、更直观的工作流中构建自己的预测编码模型。This new functionality will let organizations build their own predictive coding models in an easier and more intuitive workflow.

为准备即将停用,我们建议使用相关性模块的组织通过运行所有现有模型的批量计算,在停用日期之前导出其模型的输出。To prepare for this upcoming retirement, we recommend that organizations who use the Relevance module export their model’s output before the retirement date by running a Batch calculation for all existing models. 模型的所有相关性分数将永久存储在相应的审阅集内,在导出文档时可访问。All Relevance scores from your model will be permanently stored in the corresponding review set and accessible when documents are exported. 相关性分数也会作为元数据保留在加载文件中。Relevance scores are also retained as metadata in the load file. 此外,你仍然能够基于相关性分数筛选审阅集内的内容,并有权访问相关性模型生成的所有元数据。Also, you will still be able to filter content in the review set based on relevance score and have access to all metadata produced by your Relevance models.

完成未完成的模型Complete unfinished models

对于任何未完成的相关性模型,请完成评估、培训和批量计算,以便可以将模型应用于审阅集内的文档。For any unfinished Relevance models, please complete assessment, training, and Batch calculation so that you can apply the model to the documents in a review set. 完成批计算将在相关性模块停用日期后保留信息。Completing the Batch calculation will preserve the information after the retirement date of the Relevance module.

以下是完成任何未完成模型的步骤:Here are the steps to complete any unfinished models:

  1. 对模型进行训练,直到其稳定下来并准备好进行批计算。Train your model until it is stabilized and ready for Batch calculation. 请参阅 标记和相关性培训See Tagging and Relevance training.

    以下屏幕截图显示了一个可供批量计算的模块。The following screenshot shows a module that is ready for a Batch calculation. 请注意,评估和培训已完成,下一步是运行批计算。Notice that the Assessment and Training is complete, and the next step is to run Batch calculation.

    可供批量计算的模型屏幕截图

  2. 运行批处理计算。Run the Batch calculation. 请参阅"执行批处理计算"。See Performing Batch calculation.

  3. 验证批计算是否成功。Verify that Batch calculation was successful. 请参阅 批计算结果See Batch calculation results.

有关完成未完成的相关性模型的帮助,请联系 Microsoft 支持部门。For help with completing unfinished Relevance models, contact Microsoft Support.