Sentiment analysis is derived from machine learning algorithms and is typically used in recommender systems in order to suggest what books or movies that you might like. It is based on your personal preferences and reviews by users.
For project management this can be applied in a number of ways:
1) An analysis of documents such as a scope statement can be done in order to determine if it is likely to result in a positive or negative project outcome. The method is to assess scope statements for successful projects and compare them to scope statements that have caused significant project problems. While it may seem obvious to an experienced project manager who identifies gaps in a document, a machine learning algorithm has superior capability. How much detail from previous projects and documents can a project manager remember? A machine learning algorithm can input and analyze hundreds or even thousands of previous projects and is able to identify key elements that are likely to create project problems.
2) An analysis of organizational emails and instant messages can be done in order to evaluate if project sentiment is favorable or unfavorable. This can be used as a feedback mechanism to evaluate the effectiveness of project communication.
3) An analysis of organizational emails and instant messages can be done in order to verify or change actions used to manage project stakeholders. This is certainly an area of controversy with possible privacy concerns. However, this is about what is possible not what may or may not be ethical. An organization may decide that internal computer systems are open for AI software programs. A sentiment analysis tool can detect changes in the level stakeholder engagement and recommend corrective actions. This can provide valuable information for a project manager who receives early feedback that a stakeholder who was positive is now turned resistant to project implementation.
These are only a few of the possibilities for sentiment analysis in project management. Availability of these tools is imminent and organizations will be challenged to decide how they will be used.