个性洞察 - 分析来自不同用户的评论
Personality Insight - Analyzing comments from different users
我正在发送一个 JSON 文件，其中包含来自不同用户的多条评论和评论，但似乎回复一般，而不是针对我通过 "content" 字段发送的每条评论JSON。 API 是否能够 return 洞察每个用户的每条评论？还是将其视为单个用户的多条评论？
例如，我有不同用户的评论：
{
"contentItems": [
{
"content": "amazon's business model is incredible, is riding a number of secular trends (ecommerce, cloud, ai), and the stock is a winner. employees are making more money than expected.+ leadership principles drive a high performance culture that focuses on customers. it feels great to work on products that customers love.+ you get to work on very difficult problems with smart people. once you establish yourself as a high performer, you have a high level of job security and internal mobility. teams are constantly hiring and building really innovative things and you are encouraged to move around and explore.+ teams tend to be lean and you will be asked to learn a lot quickly. ownership is highly valued.+ office environment is really desirable. located in a great downtown seattle neighborhood, many people walk to work, bring dogs to the office, and restaurants and bars are very accessible.+ amazon veterans tend to be incredibly talented individuals, and other companies realize it. being successful at amazon is well respected in the industry.. - work/life balance can be a challenge. work demands are high and teams are often too lean. you have to set your own boundaries. even with kind managers, overachievers will feel under water.- frugality as a core value goes overboard. if amazon doesn't have to give it you, it won't. no perks, no free food or drinks, bad coffee, unsubsidized cafeterias, mediocre hardware for non-technical people. there doesn't seem to be a morale budget and you will have few official team outings.- compensation policies are not employee friendly: 401k matching is subpar. once your signing cash bonus is fully vested, your entire compensation will be base salary and stock. base salary is capped at ~0k across the company. stock vests twice a year if you're below a director, so your compensation is very lumpy. stock price appreciation is taken into consideration in your total compensation targets (ie if the value of previously offered shares increases, the company will count that as a raise and might not grant you additional stock bonuses, despite strong performance).",
"contenttype": "text/plain",
"id": "19"
},
{
"content": "--> culture & leadership principles - the amazon leadership principles are not a mission statement that was developed and then forgotten. we live and die by these principles and they drive our culture... and the culture is pretty great!--> work with super smart, very talented people",
"contenttype": "text/plain",
"id": "250"
},
{
"content": "1) when christmas season is busy time, management seems nice to everyone and act like they love and respect that you came to work each day, you can do no wrong as long as your busy. after a year i now know that working for wal*mart just sucks monkey balls!!!!!",
"contenttype": "text/plain",
"id": "66674"
}
]
}
我希望每条评论都能输出带有个性见解的输出。相反，我对所有评论都有一个一般性的回应。所以，我想知道该服务是否能够做到这一点，或者我是否必须为每个用户发送一个 JSON？
该问题与 This 其他问题相关，但我需要处理来自不同用户的 1000 条或更多评论，而不是单个用户的 1000 条推文。
您需要向每个人发送一个请求，但即使可以向多个用户发送一个请求，为每个用户获取一份配置文件，在您的示例中，两个内容将 return 作为一个错误。那是因为PI至少需要100个字才能给出结果，而最后两条推文不到100个字，所以你将无法获得个人资料。
This table 可以为您提供更多有关获得个人资料所需的字数以及使用少量字数的缺点的更多信息。
如果在您的用例中通常不可能获得至少 100 个单词，则使用 PI 可能不可行，但您至少可以获得一些有关用户语气的信息，with Tone Analyzer，这不是一回事，但可以提供有关每个用户的一些信息。
我正在发送一个 JSON 文件，其中包含来自不同用户的多条评论和评论，但似乎回复一般，而不是针对我通过 "content" 字段发送的每条评论JSON。 API 是否能够 return 洞察每个用户的每条评论？还是将其视为单个用户的多条评论？ 例如，我有不同用户的评论：
{
"contentItems": [
{
"content": "amazon's business model is incredible, is riding a number of secular trends (ecommerce, cloud, ai), and the stock is a winner. employees are making more money than expected.+ leadership principles drive a high performance culture that focuses on customers. it feels great to work on products that customers love.+ you get to work on very difficult problems with smart people. once you establish yourself as a high performer, you have a high level of job security and internal mobility. teams are constantly hiring and building really innovative things and you are encouraged to move around and explore.+ teams tend to be lean and you will be asked to learn a lot quickly. ownership is highly valued.+ office environment is really desirable. located in a great downtown seattle neighborhood, many people walk to work, bring dogs to the office, and restaurants and bars are very accessible.+ amazon veterans tend to be incredibly talented individuals, and other companies realize it. being successful at amazon is well respected in the industry.. - work/life balance can be a challenge. work demands are high and teams are often too lean. you have to set your own boundaries. even with kind managers, overachievers will feel under water.- frugality as a core value goes overboard. if amazon doesn't have to give it you, it won't. no perks, no free food or drinks, bad coffee, unsubsidized cafeterias, mediocre hardware for non-technical people. there doesn't seem to be a morale budget and you will have few official team outings.- compensation policies are not employee friendly: 401k matching is subpar. once your signing cash bonus is fully vested, your entire compensation will be base salary and stock. base salary is capped at ~0k across the company. stock vests twice a year if you're below a director, so your compensation is very lumpy. stock price appreciation is taken into consideration in your total compensation targets (ie if the value of previously offered shares increases, the company will count that as a raise and might not grant you additional stock bonuses, despite strong performance).",
"contenttype": "text/plain",
"id": "19"
},
{
"content": "--> culture & leadership principles - the amazon leadership principles are not a mission statement that was developed and then forgotten. we live and die by these principles and they drive our culture... and the culture is pretty great!--> work with super smart, very talented people",
"contenttype": "text/plain",
"id": "250"
},
{
"content": "1) when christmas season is busy time, management seems nice to everyone and act like they love and respect that you came to work each day, you can do no wrong as long as your busy. after a year i now know that working for wal*mart just sucks monkey balls!!!!!",
"contenttype": "text/plain",
"id": "66674"
}
]
}
我希望每条评论都能输出带有个性见解的输出。相反，我对所有评论都有一个一般性的回应。所以，我想知道该服务是否能够做到这一点，或者我是否必须为每个用户发送一个 JSON？ 该问题与 This 其他问题相关，但我需要处理来自不同用户的 1000 条或更多评论，而不是单个用户的 1000 条推文。
您需要向每个人发送一个请求，但即使可以向多个用户发送一个请求，为每个用户获取一份配置文件，在您的示例中，两个内容将 return 作为一个错误。那是因为PI至少需要100个字才能给出结果，而最后两条推文不到100个字，所以你将无法获得个人资料。
This table 可以为您提供更多有关获得个人资料所需的字数以及使用少量字数的缺点的更多信息。
如果在您的用例中通常不可能获得至少 100 个单词，则使用 PI 可能不可行，但您至少可以获得一些有关用户语气的信息，with Tone Analyzer，这不是一回事，但可以提供有关每个用户的一些信息。