At 9 a.m. on October 8, Monday, the Institute of Natural Language Recognition, Comprehensive Experimental Building, China Overseas University of Technology.

The warm sunlight shone through the screen window on the sandalwood-colored desk, leaving patches of light and dark alternately, and Professor Tang Tianhua, wearing reading glasses, was holding a pen to revise the material.

At this moment, there was a knock on the office door.

"Come in. ”

The door was unscrewed open from the outside with a "click", Professor Tang subconsciously looked over, and when he saw the person who came in, there was a smile on his unsmiling face.

The person who came in was naturally Chen Xu, and he walked to the desk and shouted a little crampedly: "Hello teacher!"

Professor Tang smiled and stretched out his hand and gestured, "Sit down and say." ”

Chen Xu walked to the solid wood chair at the base of the wall and sat down.

Professor Tang moved his body, leaned back in his chair and asked with a kind smile, "How is it, have you decided on a research direction?"

To be honest, Professor Tang also had a good understanding of Chen Xu's homework, and when he saw those transcripts that were not even excellent, he was a little shaken in his heart, thinking about whether to observe and observe?

But when the paper amazed him, all doubts were dispelled.

This is a young man who is very talented in natural language processing, and he plans to train it well.

Chen Xu scratched his head and said, "Uh...... Lexical and syntactic analysis. ”

Hearing Chen Xu's choice, Professor Tang frowned.

As can be seen from the article "Prospects of Natural Language Processing", this is a young man with great ambitions, and it is reasonable that he wants to challenge the core problems of natural language processing in the world today.

But well......

Professor Tang said: "You have to think clearly, once you choose the wrong direction, you may take a lot of detours, waste time and energy, are you sure to choose lexical and syntactic analysis?"

In front of Professor Tang's persuasive words, Chen Xu nodded firmly, "I'm sure." ”

No. 6 South Road, Zhongguancun Academy of Sciences, HD District, Yanjing, is the headquarters of the "China Computer Association", and it is also the office location of the top computer journal "Journal of Computer Science and Technology" and the English name JCST.

At 9 o'clock in the morning, Professor Liu Chenglin came to the association to handle some official business.

Liu Chenglin is not only the vice president of the China Computer Association, the head of JSCT in China, and the authoritative professor of speech recognition in China, but also a doctoral supervisor at Aoki University.

At half past 10 o'clock, after Liu Chenglin finished dealing with the backlog of official business during the National Day, he came to his office to take a short rest, and then prepared to go downstairs for lunch.

I walked to my desk and glanced at the desk, and sure enough, there was a thick stack of paper manuscripts.

Liu Chenglin is used to this.

Now artificial intelligence is a popular major in computer science, and speech recognition is a hot topic, so many colleges and universities across the country, you can imagine how many people are engaged in related research?

As an authoritative expert professor in this field, he receives many requests to help review manuscripts every day. They are all students and juniors of some academic colleagues and friends.

Taking advantage of the moments before lunch, Liu Chenglin wanted to see if there were any novel ideas in these papers.

Pick up the first one, titled "Development of Multi-Person Speech Recognition Software for Mobile Phone Based on Android System".

Flip it open and take a casual look, there is no novelty, no bright spots, and it does not help the improvement of existing speech recognition technology, and the only thing worth affirming is the fairly strict algorithm structure.

Liu Chenglin shook his head, put the paper aside, and picked up the next one, "Speech Recognition Based on Deep Learning and Its Application"

I opened it and looked at it, it was not as good as the previous one, it was all some fried cold rice things, and even the language expression was a little upside down, I don't know how such a paper dared to be brought to him for review.

"Who are these people!" Liu Chenglin was a little angry and threw the paper on the desk.

After thinking about it, he picked up the paper again and looked at the second author, and found that it was a professor from a second-rate university in China who he met at an international academic conference.

Liu Chenglin shook his head, picked up the red pen and drew a × on the paper, and then drew a circle on the outside of the ×, indicating that the manuscript would be rejected and no longer accepted.

After calming down, Liu Chenglin reached out and took another paper from the desk.

What he didn't expect was that the paper was very thick, roughly estimated to be at least a hundred pages.

"Huh~ good guys. Liu Chenglin picked up the paper and looked at the title in front of him, and finally it was no longer "based on so-and-so", but "Prospects for Natural Language Processing".

"Yo, it's not a small tone!"

Seeing this topic, Liu Chenglin and Tang Tianhua's reactions were similar, and they wanted to laugh at the first time, coupled with such a thick stack, he suspected that this author was afraid that he was not writing the paper as a novel.

glanced at the author's name behind the title of the paper, the first author did not know, and the second author's name was Tang Tianhua.

Liu Chenglin was stunned for a moment before he remembered who Tang Tianhua was.

The two sides met at an expert meeting in China, and I had the impression that this Tang Tianhua had made great achievements in language recognition and was a rare practical talent.

However, a good impression of Tang Tianhua can't change Liu Chenglin's mood that he doesn't even have the desire to open the paper and take a look.

Barely turn the first page.

[Abstract:On the basis of identifying the 97 most frequently used keywords in the field of international natural language processing, this paper discusses the current situation and future development of international natural language processing by using the co-word analysis method, using SPSS data analysis software as a tool, through the methods of factor analysis and cluster analysis, and drawing on the latest research results.]

Although the beginning of the flat and stable beginning did not make Liu Chenglin feel bright, he also had a sense of curiosity about what the other party was going to discuss next, and wanted to see how the other party "made up"?

But Liu Chenglin's thoughts soon disappeared.

The author of this paper not only did not "edit", but everything spoke with the most authoritative data in the world.

Starting from the most basic and important lexical, grammatical, and semantic aspects of natural language processing, we will use the simplest and most unpretentious sentences to tell readers what is lexical, grammatical, and semantic analysis.

Where are the top international celebrities researching now?

What problems did they encounter?

If you're going to do research in this area, where should you start?

After reading the first paragraph, Liu Chenglin was a little amazed.

The author's essay is clear, well-written, hierarchical, precise in its opinions, and its narrative style has reached the point of returning to the basics, and he has not seen such a wonderful popular science paper for a long, long time.

More importantly, the other party has a solid foundation in natural language processing, which is something that ordinary people can't do.

Liu Chenglin couldn't help but put together the paper and looked at the name: Chen Xu.

Then Liu Chenglin opened the paper and continued to look down......

……

After returning from the comprehensive laboratory building, Chen Xu was busy.

Collect all the books on lexical, grammatical, and semantic analysis on the market, and then enter them into Xiaobai's database one by one.

Xiaobai is powerful, and does not need to tell it what is word segmentation, part of speech, granularity of words, position of the word in the sentence, synonyms, transliterations, etc., etc.

However, in the future, the translation software/machine developed based on natural language processing will be put into commercial applications, and it is impossible for Xiaobai to act as a smart brain to translate, which is also unrealistic.

Therefore, it is necessary to overcome the problems in natural language processing.

For example, how can machines understand human emotional, ambiguous semantic expressions?

This requires Xiaobai to keep deducing on the basis of today's natural language processing research and crack the ultimate code of lexical, grammatical, and semantic analysis.

When the time comes, it will be easy to carry out syntactic analysis, text analysis, and the establishment of language cognitive models, etc., and the entire problem in natural language processing will fall to the ground like dominoes.

And he, as Xiaobai's "life mentor", needs to lead Xiaobai to work hard in a certain direction, just like constantly filling in the writing material library before.

In addition, in the future of software development, he will have to communicate with experts and scholars from all over the world as a "developer", and it would be too unreasonable if he did not even understand the basic principles of technology.

Therefore, while Xiaobai is learning, Chen Xu also started reading popular science books on artificial intelligence, in order to help Xiaobai grow better......