Skip to main content

Week 5 [23.04-29.04.2018] - How tough is it to make computers understand us?


NLP (Natural Language Processing)

One of the topics which have been a quite important subject of studies for quite a long time is computer’s capability of understanding us. In a way, it always did. So far, it was only a matter of using a predefined commands, interfaces or using programming languages to give precise instructions and get equally precise outcome as a result. This topic focuses on a more specific  interaction between humans and computers. Natural language processing is an area of computer science and AI dedicated to researching methods of enabling the machines to process natural language data (the languages that humans speak.)



Questions:

  1. How often do you use solutions of NLP? Which ones? Can you imagine living without them?
  2. Do you consider this topic as an important area of computer science? Why / why not?
  3. Do you consider this technology (or at least some parts of it) mature? If not, how long will it take to develop it to be?

Comments

Jakub Nietupski said…
I use Google Translate quite often and sometimes I turn on automatically generated subtitles on YouTube. Speech recognition is not strictly NLP but it's a related topic as well. I can imagine living without them, but they are definitely useful.
In my opinion it is an important direction of development because it creates new possibilities for humans to interact with the machine. For example it is possible that NLP will help people with disabilities to use computers.
NLP is definitely developing fast with the rise of neural networks and because it is such a mysterious branch it's hard to predict how long exactly it will take it to fully develop but it is likely that we will achieve very good results.
Lukasz Mroczek said…
I’m not pretty sure but I suppose that I sometimes use some solutions of NLP. To be honest I only use them for academic purposes like modeling processes etc. That’s why it’s almost obvious that I can imagine living without them. In my opinion, this is not the very important topic for computer science but sometimes it’s very helpful and allows us to understand some processes. I think this technology is mature but it maybe not developed enough.
Unknown said…
I very rarely use NLP solutions (for example Google TTS engine). Although NLP technology sometimes comes in handy, it still lacks precision and is very limited in its capabilities.
I think it is an important topic in Computer Science because NLP will allow us to interact with electronic devices in more convenient and natural way.
For now, I don't think that this technology is very mature. Like I mentioned before it has limited functionality. Even using machine learning approach, I think it would take lots of time for us to perfect this technology and adapt it to everyday use.
Jakub Lisicki said…
I couldn't agree more with the statement that this area of development is useful. The hypothesis that it could help the people with various disabilities is not a hypothesis anymore. Thanks to the neural networks responsible for understanding human speech, there is no need in typing any text anymore (at leasts in some aspects.) Such people can finally use modern solutions for checking the weather, making notes, writing/reading e-mails using voice assistants, etc. Speech recognition is not strictly NLP, but it's one of the main topics which are connected to this area. I think that most of the modern solutions of natural language processing rely on that completely or at least to a certain extent.
Jakub Lisicki said…
I'm definitely not criticising anything in your statement, but I recommend you to take a little more time into research of the subject. Personally, I think that this topic is considered as the future of the computer science. It's not that easy to overlook, too! Just look at the several examples on the market, among them: Siri, Bixby, Cortana, automatically generated subtitles, Google Translate functionalities for translating speech to text. There is still really much to do in this topic, but with such interest we can achieve really great things in the future!
Jakub Lisicki said…
Sadly, I can agree that NLP technology still lacks precision and is pretty limited. Someday it would be really great in terms of human-computer interaction, but for now it's mostly used for fun and some basic assistance applications. The thing with involving machine learning approach is that it's both extremely expensive to gather enough data to make it precise and it takes way too much time and effort for it for the process of development itself, to be called mature. It's already adapted to everyday use, but only to some extent. We can't expect it to be human-like anytime soon.
Anna Koca said…
I do not use the NLP technology very often. None of my friends uses Siri, Bixby, or other market-leading smartphone voice assistants, as they are still full of errors, and instead of shortening the way to get a function executed, they make it take longer.
Of course, it is undoubtedly the future of devices and technology, but still requires years of improvement. Some of us have seen the movie "Her", where the NLP system is so advanced that it begins to have "own feelings" and also provokes feelings in its user. Other movies show some threats of advanced AI - such as androids taking over the Earth and annihilating humanity, and so on.
So, like I said, for now NLP is neither immature, nor mature, in my opinion. For it to be truly useful and to actually make user experience better, it will still take a couple of year more.
Sometimes when I drive a car, I use the Google Maps option, thanks to which I can enter my destination without typing. The biggest difficulty with such technologies is that each language is different. I believe that the language processing algorithms are still underdeveloped. I think that applications based on NLP can be helpful for elders who can not use computers and phones. In addition, it's quite convenient when you have your hands occupied and you need to check something.
Unknown said…
I often use. For example, I use automatically generated YouTube subtitles. I can imagine, I think that if it disappeared, nothing bad would happen.

I think it's important. It is worth developing this technology because it can facilitate our duties or pleasure.

Tough question. Of course, at the moment it is already developed and helps in everyday life. However, I think that this is still the beginning and technology should develop all the time. We will see how be in the future, because I am curious about the technology.
I am not a big fan of NLP solutions, I mostly use it when I am bored or I want to measure its accuracy. It’s pretty easy for me to imagine living without such advanced helpers. In my point of view NLP solutions are waste of pretty huge amount of resources and power that gives not so much. I still prefer using default keyboard over Siri.

I consider this topic to be rather fun and challenging than important. As I said it demands a lot of knowledge but produces not so much profit. It is something that simplifies our lives a bit but it could be easily ommited.
Andrzej Gulak said…
This comment has been removed by the author.
Maciej Główka said…
To be honest, I seldom use NLP solutions. I would probably use them more often, if they would support polish language.
I think it is a very important area. it helps many people, especially older people, in operating computers, mobile phones etc. That's why it is important to make it better and better.
Is it mature technology? In my opinion english and other common languages versions are fairly mature. I hope that one day there will be such mature polish version.
Unknown said…
I'm not a regular user. I use YouTube subtitles and speech to text recognition when using maps.

NLP itself is very interesting but in terms of importance I think it is got a very high priority. However, in my opinion NLP could be leveraged along with whole AI technologies to craft a brand new interface to communicate with computers finally getting rid of keyboards, but this is very far away if not unreachable.


Some time ago I stumbled upon a functioning product. It is an interface for databases. I have not tried it and I am giving the link only for the sake of sharing not advertising - https://friendlydata.io/. It's just interesting.
Vladlen Kyselov said…
This area is important for us, no doubt. To my mind, it is not developed enough for now so we have to wait a little bit more to get a decent results from NLP algorythms. I don`t think it will ever become a mature language, perhaps for machines only in the nearest future.
Illia Shynder said…
I think I have never used it. Like, i just don't need it in my life. Or i used it once or few times just to test. And as my experience wasn't that good, I doesn't want to use it any more. So my thoughts cant be reliable. But i think, that it's not the most important nor even important part of development. I think, that this topic may be developed and we can make something bigger and far more interesting and advanced, than just Siri or tools like it. And yeah, I think that these technology and topic are raw yet, and need further development and investigation.
Natural Language Processing is good thing in case when we need to communicate with machine because this allows for us without extra knowledge control some things. At this time we have such well known NLP systems like Cortana, Google Now, Siri, Bixby etc. this systems simplify for us some tasks because we can tell something like "Set alarm to 10:00" and that's all that need to do to set alarm.
Unknown said…
1. If I already use any of the solutions of NLP, it's usually Google translator, because it's fast to use. I think I would survive if it was missing, there are always paper dictionaries.
2. I think this is an important topic for further development of computer science. It is important that this continues to develop, as it helps to better develop human-computer interaction
3.I think there is quite a developed technology that helps in everyday life, but if you can call it mature, I don't know. Surely it is closer than ever and not much missing.
Marcin Górski said…
To be honest, I have never used Natural Language Processing in normal situation. Of course, I have tested it but only that. I'm not so patient and I'd rather to type destination in maps application once than repeat x times the same words because device can't understand me.
Yes, it's very important because it can make things easier for us. I think that NLP still has a lot of bugs but developers have a lot of time to improve that solution.
Filip Sawicki said…
People forget that NLP is not only a tool for voice assistants, but has a great role in text analysis and prediction. Press, banks, financial and statistical institutions, governments and lots of technological companies use it daily to create better solutions. I can imagine living without this technology however without it I’m totally sure that then we’d have a global crisis. Why? Well because all of hedge funds and investment banks base their market analysis on NLP. I think that NLP is a mature technology and we’ll it see more and more in the future. One think that I’d change is to focus more on other languages, we have great tools but most of them are trained on English.
Jakub Lisicki said…
Actually, that's right. Safety is another reason for developing this technology. There are so many people who say that this technology can find use only in the entertainment purposes, but that's true only to some degree. It is used mostly for entertainment or at least seen as such, but it really helps and could possibly help the disabled even more in the future.
Jakub Lisicki said…
Creating the advanced AI is the aim of many researchers for many years and I couldn't really imagine human-like AI without the ability to speak nor understand what am I saying. Of course, it's just another step of development of such an advanced technology, but it's also important. Apart from that, like I've mentioned in some of my other comments, it could really help the disabled with so many issues they have to struggle with nowadays.
Jakub Lisicki said…
As many of us know already, it's extremely expensive to train the tools of NLP on any languages, that's why they're not adjusted to most of them. Out of the ones that are the focus of this technology, you are absolutely right about the other uses of it. Many market adjustments are based on the analysis of comments for the current products on the markets. With help of such tools it becomes really easy to predict the undertone of most of the comments and estimate the impact it has on our company earnings and losses.
Unknown said…
Actually not really using any of NLP solutions. Played a bit with google assistant and bixby but usually its faster just to do whatever you want to do in old fashion finger way.
It depends on what you understand under computer science. It might be but for progress in general I feel like by itself it wouldn't be of much help.
Well, scientist always say that in few years we will have such a technology and this started from 90's so my save guess would be something like 2025 to get somewhat human like interaction with computer
Unknown said…
1. I don’t use voice recognition at all. However, if it were to improve significantly along with language translation I imagine I would be using it a lot.
2. Definitely, as stated in the video. This form of machine interface is more likely to be more important in the near future. Especially as normal conventional interfaces may be getting too cumbersome and complicated with time as the systems that they control get more advanced.
3. Mature as in its prime? No, the technology only made some significant advances recently due to progress in field of AI.
Unknown said…
1. I use Google Maps voice search commands on daily basis. When I am driving, they allow me to easily find and navigate to certain places without having to type their names on my smartphone. I have also played with Amazon Echo Dot for few months and I was impressed by Amazon Alexa smart assistant capabilities. It was possible to ask it questions and having simple conversation with it.
2. As a part of Machine Learning it is obviously one of the most important areas of computer science nowadays as it provides new techniques of Human-Computer interaction and as such will be definitely used on a mass scale.
3. I consider this technology as constantly and rapidly progressing, most of the solutions are still under heavy development and I think it will take them 2-3 years to be market-ready products.
Jakub Lisicki said…
The main issue with today's AI is that it is fairly capable of having a simple conversation, but it lacks the capability of deeply understanding anything that we say. That's why Amazon's Alexa may seem to be pretty nice at first, but we could get bored of it after a while. It does have a variety of things to do with it, but the list is still limited. I hope that someday, that list would have no limits and that the conversations with the machines could actually be undistinguishable from conversations with humans.
Even though some of the areas are already fairly mature, the pace of the progress is so fast that we can call some of the 2-3 years old books on this topic "old." I don't think that's bad, it just shows how rapidly this area is evolving.
Unknown said…
How often do you use solutions of NLP? Which ones? Can you imagine living without them?
Unfortunately those technologies have no use of polish language witch make them almost useless at Poland. I do not use any of those because of that. I heard that Google works at polish version of speech recognition right now, which make it closer to us.
Do you consider this topic as an important area of computer science? Why / why not?
Yes, In my opinion it is important area, because that could lead to communicate with computer without using mouse and keyboard.
Do you consider this technology (or at least some parts of it) mature? If not, how long will it take to develop it to be?
In my opinion there are a lot to do in that area. How long? That could be 10 or 50 years.
Unknown said…
Honestly, I barely use them. I don't remember when was the last time I used some solutions of NLP. I think I only use them for academic/school purpose like some projects. I don't concider that in "very important" categories for computer science, maybe I'm not initiate in IT life too much. I have to tell thats it’s very helpful and allows us to understand some processes - but as I mentioned before, I don't work on this type of project often. In my opinion this technology is mature, but it can always be better. Time will tel how long does it take to fully develop this field of science.
Since my mobile hardware is old and does not support NLP in most cases i dont really use it. I tried using it while working on VR project which opened quite a few new ways of controlling while in VR. Ive thought about dedicating some time to setup a Google assistant on my PC and check how it works but never really went to do it.
The NLP topic is big in software engineering and machine learning, improving the way it works and how much and how fluently it can understand context is a huge step in making programming, user interfaces and global communication better.
NLP will be our ui of the future.
No not at all, although we have come very far from where we have been in as far as 2014 we are still baby stepping in darkness when it comes to fluent language recognition software that could replace whole UI's and provide ease of access.
Peter Clemenza said…
Yes I could live without some basic translators. However translators
that we are mistakenly trusting more and more.

I think there are more interesting topics to consider than this one.

This technology will be fully developed when the real AI will be developed - because of
different meanig and content of text.

That I see as I gaze so far ahead? Well generally some implants that are melted in our brains
that are going to read our thoughts.
Of course, i use them. I think everyone is using them. It's hard to imagine but i can image living without them. Everything in our world can be interchangeable. Definitely yes. This topic is so important for every user i think. So many problems we can meet when we are using PCs. But NLP can rapidly help us. It is already developed but not in a good way to represent for all the users. After 10 years we will use it everywhere.
Illia Lukisha said…
I use it not often but when i need it - i use it. Which ones? Google maps invoice is a good thing but not truly perfect. They need a lot things to do to improve it. This topic is a future and you don't need to discuss it. I don't think that this kind of technology is mature enough. It will be useful for 100% after 20-30 years.
Iman Masjedi said…

big data a solution for natural language processing, Because you can be more sure about experiments, you can do them quickly, you can apply them quickly, you can correct them quickly.

Patryk Górski said…


I use Siri, and also a google translate. It's very easy to live with them, but I think that I can live without them.

Yes. Language processing is important, it can lead to a lot of great things.

I think it is not ideal yet. There is still a lot to improve. However, I am satisfied with it's level right now
Unknown said…
Shame on me, but probably more often than I think. Most of the inputs on the internet nowadays use NPL to improve the advertisements they are showing to us. I use Siri from time to time, I hope they will grow it in the upcoming years or some good competition would show up. I think it's an important area of computer science because I see great potential in it. I hope in someday we will use your month to communicate with our house and computer. The next step will be the assistants (like Siri but different level) who will understand complex sentences, will have their temper, humour etc. but we need lots of time. This technology is mature enough to be wildly used, but with it's potential, I think we are just starting playing with it.

Popular posts from this blog

Week 12 (12.01-18.01.15) Are you an early bird or a night owl ?

Owls are nocturnal creatures. They’re wide awake at night and they sleep during the day. If this sounds like bliss to you, then, like about 20 percent of the population who find themselves most active at around 9 pm, you may fall into the same category as our feathered friend. Night owls often have difficulty waking up in the morning, and like to be up late at night.  Studies of animal behaviour indicate that being a night owl may actually be built into some people’s genes. This would explain why those late-to-bed, late-to-rise people find it so difficult to change their behaviour. The trouble for night owls is that they just have to be at places such as work and school far too early. This is when the alarm clock becomes the night owl’s most important survival tool. Experts say that one way for a night owl to beat their dependence on their alarm clocks is to sleep with the curtains open. The Theory is that if they do so, the morning sunlight will awaken them gently and natura...

Week 11 [03-09.06.2019] The problem with ecological cars emission in UK

The problem with ecological cars emission in UK Since the adoption of the European Emission Allowance Directive in the European Parliament, all car makers have tried to submit. Since 1992, the Euro I standard has been in force, which limited the emission of carbon monoxide to the atmosphere. The Euro VI standard currently applies, which limits the series of exhaust gases. These include: hydrocarbons, nitrogen and carbon oxides, and dust.   The most significant change was brought by the Euro IV standard. For the first time it introduced the limitation of nitrogen oxides, which are responsible for the harmful compounds of smog.   What is smog?   Smog consists of sulfur oxides, nitrogen and carbon. In addition, solid substances such as suspended dust (PM). Dust suspend in atmospheric aerosols may be in liquid and solid form. These can be particles of sea salt, clouds from the Sahara and artificial compounds made by people. These compounds...

Week 4 [06-12.11.2017] This is what happens when you reply to spam email.

James Veitch is a British comedian. In today’s Ted Talk James with characteristic for himself a sense of humor shows how he deals with spam emails and why responding to junk messages may be sometimes dangerous. Questions: What do you think about James’s  way of dealing with spam? Why are junk messages legal, even though it sometimes may be a fraud? Dou you have a problem with spam? How do you deal with with it?