Since we would like to avoid the basis caused by asymmetric information and synaptic pruning of past experiences, we focus on how to test the patients’ memorizing ability by giving them new information. Current literature suggests that temporal decay and chunk capacity limits are not only the properties of short-term memory but also long-term memories (Cowan, 2008). In this evaluation, we will pay our attention to building memories by providing new information to the users and can simply evaluate their abilities to recognize and form new short-term memories in both characters and numerical digits, encoding short-term memories into long-term memory, and recalling recent memories. In this way, we can avoid as much bias caused by the different subjectivity, synaptic pruning. Also, this design will not require a very expensive program to collect as much past experience and personal information of the users as they are actually meeting with the physicians so that the design is lower-cost and more acceptable for users who are very sensitive and careful to personal information and privacy. 

There are four main types of amnesia including retrograde amnesia, anterograde amnesia, transient global amnesia, and infantile amnesia. We pay more attention to the first three types of amnesia than infantile amnesia, which is a common phenomenon of forgetting the first three to five years of life. In order to better help the physicians identify the type of amnesia, we set a questionnaire that required both the patient and the patient’s family to complete as a reference to contradict with patients’ memories about these extremely important life experiences. The questionnaire includes the significant characteristics of anterograde amnesia and transient global amnesia: 1. Have you experienced a seizure-like event; 2. Are you an alcoholic? A seizure-like event is the significant causation of transient global amnesia, which the patient may not have the memory of the experience. Meanwhile, alcoholism can cause hippocampus damage, which affects the ability to form new memories. 

After filling out the questionnaire section, the evaluation will start immediately if it is in the morning, or the next day, if it is not in the morning to confirm the duration of the 12-hour test is long enough to evaluate the patient’s long-term memory. We try to quantitate the test result by giving each question certain credit for directly visualizing how well or how bad the users work on the tasks. In the morning after the users get up, the reminder function of the application will notify the user to start the evaluation tasks. First, the application will randomly select 3 words from dictionary.com with 7 to 13 alphabets and pop up on the screen of the device, and requires the user to recognize and memorize them in 10 seconds. Since the newly formed information requires 30 seconds to integrate into short-term memory. After a 30 seconds break, there will be 3 multiple choice questions with 4 possible words (all will have similar length letters) and will be asked which one is the answer shown on the tablet. The patients need to recognize and select the words that have shown up previously in each question. Each question has a maximum of 10 seconds to answer. This can test the recognition ability and short-term memory of the user. A correct answer will be given 1 point. After that, a randomly generated 10-digits number sequence will be given to the patient for 30 seconds for memorizing. Then the patient needs to recall the correct sequence of the 10-digit number. Correctly answering one digit at the correct position will gain 0.3 points, with a maximum of 3 points. This means the users can still get credits for answering the correct number for the sequence end even though some middle part of the sequence is answered incorrectly. Based on the previous study, there are cognitive bias components among controlled patients, which are context-independent, context-dependent, and mixed (Stratta, et.al., 2000). The design of mixing alphabets, numbers, and context-relevant questions aims to avoid the situation that users that are less sensitive to words or alphabets and have a bias in responses may be considered as amnesia patients. The last step of the morning part of the evaluation is giving the patient approximately 600-word head-top news. The users need to finish reading the news in 5 minutes. 

The second part of the evaluation is to evaluate the long-term memory and recalling ability. Before sleep, the users need to correctly choose the 3 words given in the morning from 10 choices. Each correct selection is one point. After selecting 3 from 10 words, We will test how well the user can recall the news they read in the morning. This test will have 3 questions, each worth 1 point and all questions will be generated through the prewritten program. The first question will ask for “which of the following is the topic of the news you read this morning?” Only one out of the four choices is correct. The answer choice will be the category of the application program chosen from the TIMES, ABC news, NBC news, or CNN news, etc. The second question will ask the participant to identify which of the articles/sources of the news they have read from. The last question will ask the participant to identify the image that was presented in the news.