At step 320, the message analyzer 232 may analyze the first communication session and the second communication session to determine similarity or time distance between the first and second communication sessions. For example, the message analyzer 232 may determine the similarity or time distance between the first and second communication sessions based on the content of the first communication session and the content of the second communication session. The message analyzer 232 may use natural language processing, machine learning, or any other technology that may process language and/or words to determine the similarity or time distance between the first and second communication sessions. For example, the message analyzer 232 may use word embedding techniques to measure similarity between linguistic items of the email messages included in the first communication session and linguistic items of the text messages included in the second communication session. If the message analyzer 232 determines that linguistic items of the email messages and linguistic items of the text messages have significant overlap (e.g., 40% of the linguistic items in the email messages are same as or similar to the linguistic items in the text messages content), the message analyzer 232 may determine that the first and second communication sessions have high similarity. As previously discussed, as another example, if the message analyzer 232 determines that the first communication session and the second communication session share a common topic, subject, or keyword, the message analyzer 232 may determine that the first and second communication sessions have high similarity.