英国硕士论文：情感或情绪模型
选择的方法将会在解释情感分析、情感模型选择和使用Python代码方面进行讨论。
从决策的文献来看，决策任务受即时情绪和预期情绪的影响。预期情绪可以被认为是用户应该证明的决策结果的影响。即时情绪是最近发生在外部并进一步影响用户的事件的结果。基于情绪感知的推荐系统应注重识别即时情绪和预测预期的情绪结果。推理的过程应该是在立即确定的情绪状态下，为某一用户产生积极的预期情绪提供建议。决策过程中的情绪可以通过显式或隐式策略来检测。明确的策略是建立在与用户直接互动的方法的基础上，获取用户在决策过程中的情绪信息。可以使用问卷来确定用户的情绪和个性特征。
再往前走，更多的时候，个体无法正确地解释他们的情绪，因此，在这种情况下，明确的策略不能被认为足以正确地识别即时情绪。由于这个特殊的原因，可以采用隐式策略。一些研究人员曾表示，可以通过视频检测情绪，但同时保持有限的准确度范围。Poria et al.(2015)提出了一个多模态框架，利用文本源、视频和音频来识别用户的情绪，并将其映射为Ekman的六种情绪。结果表明，通过组合不同的信号，可以在情绪检测任务中达到最高的精度。根据这项工作，可以使用隐式来源通过文本获取即时情绪信息，同时具体收集用户社交网络活动中的帖子。
英国硕士论文：情感或情绪模型
Methods selected will be discussed with regard to explaining sentiment analysis, emotion model choice and used Python code.
As per the literature of decision making, the task of decision making is affected by immediate emotions and expected emotion. Expected emotions can be considered as the affects that the user should be proving as a result of the decision made. Immediate emotions are the result of an event taken place externally and further affecting the user recently. Recommender systems based on emotional awareness should focus on identifying immediate emotions and forecasting expected emotional results. The process of reasoning should be providing recommendations for the generation of positively expected emotions for a certain user within an immediately defined state of emotions. Emotions during the process of decision making can be detected by the use of explicit or implicit strategies. Explicit strategies are set on the basis of method having direct interaction with the user for obtaining information about their emotions during the decision making. There can be use of questionnaires for the identification of both, user emotions and user personality traits.
Further ahead, more often, individuals fail in explicating their emotions correctly and hence, in such cases, explicit strategies cannot be considered sufficient for the correct identification of immediate emotions. Due to this particular reason, there can be an adoption of implicit strategies. It had been stated by a number of researchers that there can be detection of emotion through video, but while maintaining the scope of limited accuracy. A multi- modal framework had been presented by Poria et al. (2015) that utilize text sources, video and audio for identifying emotions of user and for mapping them as the six emotions of Ekman. The results depicted that there can be achievement of maximum precision within the task of emotion detection by the combination of various signals. In accordance with this work, there can be a use of implicit source for obtaining information on immediate emotions by the text, while specifically gathering posts from the social network activities of the user.
