Reading suggestions on an individual transcript page

Currently, each transcript contains metadata like `tags` and `speakers`, which give information about what the transcript covers and who the speaker is. We can use this information to provide suggestions to readers based on the transcript they are currently viewing.

To implement this feature, we can explore leveraging recommendation algorithms.

Moreover, extending this functionality to the Review Product can benefit reviewers by offering suggestions, thereby enhancing their workflow, productivity, and engagement.

首先定位单个 transcript 页面，并检查其标签和说话人元数据是如何公开的。定义推荐方案，并确定 Review Product 是否在范围内。完成的标准是读者能够根据 transcript 元数据获得相关建议，并且双方已就两个界面的预期行为达成一致。

由索引模型根据 Issue 内容生成。