Chapter 97 Developing a professional model

Lin Hui clicked on the specific comment.

Wow! Almost all short compositions.

But Lin Hui really wants to see what the users of this software think about it.

Although Lin Hui’s main purpose in developing Nanfeng APP is to let Internet giants notice how awesome the algorithm of this software is.

But users are not the stepping stone to Lin Hui’s success.

How should I put it? Every user is an individual with independent thoughts and feelings, not cold data under an algorithm.

Their ideas should also be taken seriously, and their reasonable suggestions should be considered and accepted.

In short, user feedback is also worth noting.

Even though each comment was quite long, Lin Hui read it very seriously.

Lin Hui clicked on the top comment in order.

“It is difficult to write a review about Nanfeng APP. Although I spent the whole day on it today, the more I use this software, the more powerful I feel...

My job requires a lot of reading/research, and Nanfeng APP saves me from being buried in endless piles of paper. Objectively speaking, this software has surpassed Yahoo News Summary in all aspects..."

Well (⊙﹏⊙), such comments are really dog-licking comments. Although there are many compliments, they have no reference value.

Lin Hui then turned to other comments:

"I am a journalist, and I like Nanfeng APP very much. Its layout is simple and intuitive, with only one input interface. It looks a lot like a commonly used translation software. Although this design is a bit unusual, I have to admit that it is

The design makes it easy for me to concentrate on my work.

I used to use Yahoo News Summary Professional Edition, but I have to say that the summary efficiency of this software is much higher than that of Yahoo News Summary.

Some areas that need improvement: First of all, I suggest being able to enable the processing of multiple news summaries at the same time, because a professional news editor often needs to process a large amount of news at the same time;

Secondly, I suggest that this software does not limit the number of words in the input file to 10,000 words, because many of the news that news editors need to process cannot actually be called news, but rather deal with interview scripts, and interview scripts are generally more than 10,000 words.

This input file limit of 10,000 words is really unfriendly.

Of course, this feature is only needed by professionals. If developers can make this improvement, I will pay for it even if it is put in the professional version, because such a feature is really important.

In fact, this software has helped me a lot now, it has greatly helped improve my work efficiency, I will stop here. Thanks to the developers for making a great app, fully looking forward to this

The future of applications.”

This comment is much more valuable than the one just now. At least through this comment Lin Hui can know the shortcomings of the software and the direction for improvement.

The most important thing is that Lin Hui knows the attitude of people in the industry towards this software.

Judging from the above reviews, journalists quite agree with this software.

As for the suggestion this person made.

Breaking through the 10,000-word material input limit is not too complicated.

In fact, the software Lin Hui originally developed can support a single input of up to 50,000 words.

It's just that I don't understand the press release.

Lin Hui thought that the input limit of 10,000 words at a time was enough!

Now it seems that it was still not well thought out.

As for processing multiple news summaries at the same time?

This is a bit unrealistic, and it involves the limitations of the algorithm.

In fact, whether it is the first generation text summary algorithm or the second and third generation summary algorithms.

When it comes to news/text summary processing, it is actually implemented one by one individually.

Mainly because although text summarization is simple, it actually takes up quite a lot of resources.

To achieve true parallel processing, at least we have to wait until the fourth generation of digest algorithms comes out.

The dependence of the fourth generation summary algorithm on the model will be greatly reduced.

However, it is basically impossible in this short time.

Now even some programs that rely on large servers claim to be able to process multiple texts at the same time.

But the actual processing is also done one by one.

Even if the user makes multiple inputs, the server will still process the first news/text first.

The first input news has been processed, but it still shows that it is being processed. The server actually starts processing the second one...

And so on, until the summary processing of the last input news is completed, and then the summary results are output together.

To the user, it may look like multiple outputs are achieved after multiple inputs.

But this kind of "multitasking" is actually just a cover-up.

Although this approach is not difficult to implement.

But Lin Hui didn't want to deceive himself.

I browsed the comment area below the Nanfeng APP for a while.

Lin Hui found that many users in the comment area strongly suggested that Lin Hui use Nanfeng APP based on

Let’s create a professional mode that is more in line with the usage habits of professionals.

It doesn't matter if you pay extra for it.

Many comments reflected this appeal.

Ah, Lin Hui has always been someone who pays great attention to user feedback.

It’s hard not to meet the reasonable requirements of users!

Lin Hui summarized the demands for professional mode from users who asked for an update in the comments.

Lin Hui was surprised to find that these people's demands were similar to the suggestions made by the guy in the comment area just now.

They all require the lifting of the 10,000-word input limit.

However, there are great differences among these people on whether to use the function of processing multiple news items.

It's outrageous to actually have a fight in the comment area.

Lin Hui took a look at the different views of the two parties.

Proponents argue that the ability to multitask increases productivity;

Opponents believe that multitasking can easily lead to distraction and affect concentration at work.

Well, as expected of a dedicated journalist, the support and opposition are entirely based on work efficiency.

Lin Hui hopes that these people will have a dispute about whether multi-tasking is necessary or not.

In this way, he can openly maintain the status quo of this controversial point.

There is no need to make a fuss about multi-tasking. Even if you want to engage in professional mode, Lin Hui's workload is much less.

Just cancel the input limit of 10,000 words.

This is not a difficult task for Lin Hui, it is just a matter of changing a line of code.

Of course, even if you just change the code, you will still be charged.

It's not difficult to change a line of code, the hard part is knowing which line to change.

The normal routine for a programmer is a cup of tea and a pack of cigarettes, and one line of code can be modified in a day.

Not to mention that one line of code needs to be modified in a day. Sometimes it is common for more than ten people to modify three or four lines of code in a month.

This is not to say that programmers are passive and slow down their work. The problem is that it is really difficult to debug the same code written by others.

I didn't play well and put the pointer around randomly.

There are no notes or anything.

The situation without notes is better. I am most afraid of encountering the kind of Martian notes that are comparable to the ghost-drawing symbols of old Chinese medicine practitioners.

It's really just one person's remarks and ten people guessing.

Why not ask the person who wrote the code?

Laughing to death, I have resigned a long time ago.

The most important thing is that when changing the code, you not only have to endure a mountain of identical code.

We also have to face constraints caused unintentionally by our peers.

Lin Hui can't say that in other industries, but the programmer industry is not suitable for large numbers of people and strength.

At least not completely applicable to the situation where there is strength in numbers.

Manpower and time do not have a linear relationship.

A large number of people cannot shorten the software development progress.
To be continued...