A Study on Giving Commonsense to Machines
0 likes36 views
1. The document discusses giving machines common sense through various techniques from artificial intelligence like machine learning and neural networks. It proposes that machines can develop basic common sense through visual processing and recognizing patterns in images similar to how humans learn as children. 2. It provides an example of enhancing a dustbin with sensors to detect objects and only open when trash is detected, displaying basic common sense. A flowchart and graph are presented to illustrate how common sense reasoning would work in machines. 3. In the future, it suggests that more household appliances and even entire homes could be automated to perform tasks based on sensing contexts like presence of people through common sense reasoning.
![International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 04 Issue: 11 | Nov -2017 www.irjet.net p-ISSN: 2395-0072
© 2017, IRJET | Impact Factor value: 6.171 | ISO 9001:2008 Certified Journal | Page 556
A Study on Giving Commonsense to Machines
Tanu.N.Prabhu1, Sashank.J2
1,2Student, Department of Computer Science, Global Academy of Technology, Bengaluru, Karnataka, India.
---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - Nowadays we are totally dependentonmachines
for each and every work. These machines are being
programmed and functioned manually, but we must allowthe
machine to think on its own, this is done by giving common
sense to machines. We must use someoftheembeddedsystems
and separate operating systems to control human gestures,
and some sensors such as microphone, VOIP sensors to record
the voice and hand gesture recognition using accelerometer
sensors. In other words whatever we tell ,the machine should
understand. In this paper we try to discuss how this concept
can be implemented to machines. This can be further
inculcated to develop fully automatic independent machines,
which tries to think on its own when given the input. This
totally works on the concept of Artificial Intelligence. There
are various aspects where this can be further developed by
designing the flowcharts. In this paper we also give some
paradigms where we can use this concept so that it can be
easy to work with machines. Basic common sense is key to
building more intelligent machines [1, 4].
Key Words: Embedded systems, Commonsense, VOIP
sensor, Accelerometer sensor, Artificial Intelligence.
1. INTRODUCTION
Artificial Intelligence is a fundamental theory and
development of computer systems able to perform tasks
normally requiring human intelligence, and thinkingsuch as
visual perception, speech recognition, face recognition,
decision-making, and translation between languages.
Nowadays the voice assistance what we use in our smart
phones like "Ok Google” mainlyworks onAIandfurthermost
the concepts such as Image recognition, Speech Recognition
also works on Artificial Intelligence. [7]
Now the next question that occurs in ourmindis howcanwe
do this, what are the strategies used in this?
And first of all how can we build machines with
‘commonsense’, the cognitive skills that every ordinary
person takes it for granted? In this paper, we describe a
common sense reasoning which is in the development stage
through understanding some rudimentary flowcharts and
some transition states of the flowchart. The system reasons
the kinds of fundamental entities that show up in nearly all
situations such as individuals, objects, occasions, objectives,
plans, and slip-ups. We first describe the main features of
our idea and then discuss its applicationandevaluationto an
artificial life scenario. [4]
1.1 Existing
This can be done with the help of Machine Learning where
the computers are fed with the inputs, can begin, on its own,
to see patterns and leads in these information,andfigure out
how to perceive and arrange new information as they arrive
into the patterns and rules that the computer has already
created. After this these data are piled up and stored in a
large database of a computer having its own rules and
regulations making it more dependable [8].
1.2 Proposed
It our Idea as to how new types of future machines can be
developed as shown below. Where we have briefed some
ways how this concepts can be implemented to the real time
application.
Some of the paradigms of automatic common sensing
machines are as follows.
The first scenario is in our rooms where the moment we
return from our work we will be sweaty or shivering of cold
due to outside temperature and of courseoneofthese.There
are some automatic AC’s which are hungonthewall whereit
will detect our body’s humidity and set the AC according to
our body’s temperature. Some thermal sensors are attached
to that so in order to detect our body temperature. If in case
our body temperature is low, then AC is set to the little more
than the room temperature which would be suitable to his
body temperature and opposite in the case when the body
temperature is high. Secondly what we are working on right
now is the common sense.
Let’s try to induce common sense to Dustbin and make even
that to have a sense too.
2. HOW DOES IT WORK?
Well a lot of common sense is requiredthough, initiallythere
are way too many dustbins in the world but our dustbin is a
way too ahead.
In the pas
followed. For instance we have to bring our hand near it, the
t days when we wanted to dump anything we used
to go near the dustbins open it up and then we used to drop
the waste there, but still as of now we still have tomove near
the dustbin. But we induce common sense to it, then only if
we move near the dustbin the dustbin should not open if in
case if we have something in our hand it will sense it and
opens .But there is an important thing here now, the dustbin
thinks that whatever is in your hand as garbage so it blindly
opens for everything there are some strategies to be](https://image.slidesharecdn.com/irjet-v4i1196-171208092457/85/A-Study-on-Giving-Commonsense-to-Machines-1-320.jpg)
![International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 04 Issue: 11 | Nov -2017 www.irjet.net p-ISSN: 2395-0072
© 2017, IRJET | Impact Factor value: 6.171 | ISO 9001:2008 Certified Journal | Page 557
point is what’s there in your hands do matter a lot. Whether
good or bad things matters a lot. Choiceiswhetheryou want
to throw it or not.
Figure 1. Advanced automatic Dustbin
A brief description about the diagram. As shown in the
Figure 1. Well, it consists of a well equipped dustbin with
embedded sensors to detect the input. And a lid on top, it’s
quite obvious and depends on the user’s choice whether it is
needed or not .In this case we use a lid just to make things
simple. We don’t have to perform some extra things
nevertheless just keep the hand above it ,There is no need of
even touching it to be more precise its commonsense and it
has a sensor.[3].
2. 1 FLOWCHART AND DESIGN.
The Flowchart Below depicts the following working of any
machine which works on the basis of common sense. Here
including start and stop there are additional of 7 states as
follows
The States are
1. Giving Inputs to the machines: This state involves various
types of inputs given by the user to the machine be it the
Hand gestures, through voice modulated commands and
many more…
2. Checks for the Valid or Invalid one: This state involves
evaluating the inputs whether is it a valid one or not. This
further consists of.
2a.If at all the input is valid then it recognises the speech via
speech or face recognition
2b.If in case the input is invalid one then it displays a
suitable message stating that it is working on it.
3. Matches the input command and checks it in the stored
database: This phase as the name suggests checks the
command whether is it stored in the database and performs
the operations according to it.
4. If the Command matches: then it performs two cases
further.
4a. If matches: Then it performs the desired operations that
meet the user requirements.
4b. If No: Then it returns to the case 2b and displays the
suitable message.
Most of the people think is it required. Whetherthecommon
sense must be given to the machines or not or someareeven
confused. Well the answer is now we are Presently in the
midst of 2017 and as we can see slowly the things are
becoming automotive such as the robots painting the car in
the assembly lane, Drilling etc. Robots in the present have
been a great asset because they canperformjobsthataretoo
risky for humans. [6]
Later on benefit robots are relied upon to outperform the
mechanical Robot showcase. Numerous makers are
beginning to investigate fields that are inadequate with
regards to specialists. One zone of intrigue is creating
automated aides for the elderly. You may likewise observe
real employment of apply autonomy in law authorization
and security. Future robots could likewise be as little as a PC
chip and embedded into apparatuses, for example, vacuum
cleaners.
If this further continues then one day everything in this
world would be automotive like automatic cooking machine
the moment you return to home from work itwill senseyour
hunger and prepare a dish for you, Ambient Light system is
when u enter into the room automatically the lights will be
turned on or off .This is how in further these machines will
be dominating the world.](https://image.slidesharecdn.com/irjet-v4i1196-171208092457/85/A-Study-on-Giving-Commonsense-to-Machines-2-320.jpg)
![International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 04 Issue: 11 | Nov -2017 www.irjet.net p-ISSN: 2395-0072
© 2017, IRJET | Impact Factor value: 6.171 | ISO 9001:2008 Certified Journal | Page 558
2. 2 GRAPHICAL REPRESENTATION.
As shown in the graph the machine is a lot more ahead than
the human .It is so believed that when we teach a machine to
perform some operations the better results can be obtained
compared to that of a human being on the graph it takes a
lead, and even the quality of the product developed by the
machine would be a lot more better than human being.
So overall teaching a machine gives a lot more sense that
teaching a human .The reason why human beings are losing
the race because they become lethargic, and often feel bored
when doing the same work back and forth. Due to this the
efficiency decreases, but on the other hand the machine
never feels bored unless and until the work is finished it
keeps on doing its work and surprisingly it never takes a
break unless it has to do so. And the speed matters a lot
because the machine works at very great speed when given
the input compared to man.
3. ADVANTAGES
It makes life simple by carrying all the tasks
It is very user friendly and easy to use in terms of
all aspects.
Communicating with humans in a simpler way.
4. DIS-ADVANTAGES
It is a bit cost effective.
Understanding the working of the machine.
Let me give an example “When you are conceived, you don't
know anything.” But as the days go on and on we try to learn
some rudimentary things through cognitive thinking, [5].
Actions which our parents do, and also through their
conversation. In the same way the machine must be built
first and there in the development stages through image
processing it will try to identifytheimagesandtheirpurpose
so when the user provides some input at the same time the
machines tries to think which image is what and what must
be done next as shown in Figure 3. So there will be
something known as visual cortex which is one of the
important Aspect in the machine which provides the ability
to think. Where there are some preliminary steps or you can
also it as an Algorithm. [2, 3]
1. Recognize the input
2. Process the input
3. React to the input
4. Perform operations according to the necessary
Steps.
5. CONCLUSION
The proposed system can be used to achieve simple and
tedious tasks through normal human interaction. The
commonsense to machine is growing to such to such an
extent that even Face book; Microsoft is trying to implement
it. As indicated by the current review Face book AI look into
a mass chief Yann LeCun, jumps forward in neural systems
administration and machine vision might one be able to day
prompt programming with the presence of mind. Neural
systems are simulated frameworks which imitate the
structure of the human mind and by consolidating this with
further developed machine vision which are approaches to
pull information from symbolism for use in undertakings
and basic leadership - LeCun says good judgment will be the
outcome. For instance, on the off chance that you have a
prevailing article in a picture, and enough information in
protest classes, machines can perceive particular items like
mutts, plants, or autos. But I promise this field has a
promising future at great heights.
REFERENCES
[1]Davis E: ‘Representations of common sense knowledge’,
San Mateo, CA, Morgan Kaufmann (1990).
[2] Minsky, M.: The Society of Mind. Simon and Schuster,
New York (1986)
[3] Cambria, E., Hussain, A., Havasi, C., Eckl, C., Munro, J.:
Towards Crowd ValidationoftheUKNational HealthService.
In: ACM WebSci, Raleigh (2010)
[4] Liu H., Singh, P.: Concept Net: a Practical Common Sense
Reasoning Toolkit.BT Technology Journal 22(4), 211–226
(2004)
[5]Speer,R.,Havasi,C.,Lieberman,H.:AnalogySpace:Reducing
the Dimensionality of Common Sense Knowledge. In: AAAI
(2008)
[6] Havasi, C., Speer, R., Pustejovsky, J., Lieberman,H.:Digital
Intuition: Applying Common Sense Using Dimensionality
Reduction. IEEE Intelligent Systems 24(4), 24–35 (2009)
[7]Singh, P.: The Open Mind Common Sense Project.
KurzweilAI.net (2002).
[8] McCarthy, J.: Programs with Common Sense. In:
Teddington Conference on the Mechanization of Thought
Processes (1959).](https://image.slidesharecdn.com/irjet-v4i1196-171208092457/85/A-Study-on-Giving-Commonsense-to-Machines-3-320.jpg)
A Study on Giving Commonsense to Machines
- 1. International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056 Volume: 04 Issue: 11 | Nov -2017 www.irjet.net p-ISSN: 2395-0072 © 2017, IRJET | Impact Factor value: 6.171 | ISO 9001:2008 Certified Journal | Page 556 A Study on Giving Commonsense to Machines Tanu.N.Prabhu1, Sashank.J2 1,2Student, Department of Computer Science, Global Academy of Technology, Bengaluru, Karnataka, India. ---------------------------------------------------------------------***--------------------------------------------------------------------- Abstract - Nowadays we are totally dependentonmachines for each and every work. These machines are being programmed and functioned manually, but we must allowthe machine to think on its own, this is done by giving common sense to machines. We must use someoftheembeddedsystems and separate operating systems to control human gestures, and some sensors such as microphone, VOIP sensors to record the voice and hand gesture recognition using accelerometer sensors. In other words whatever we tell ,the machine should understand. In this paper we try to discuss how this concept can be implemented to machines. This can be further inculcated to develop fully automatic independent machines, which tries to think on its own when given the input. This totally works on the concept of Artificial Intelligence. There are various aspects where this can be further developed by designing the flowcharts. In this paper we also give some paradigms where we can use this concept so that it can be easy to work with machines. Basic common sense is key to building more intelligent machines [1, 4]. Key Words: Embedded systems, Commonsense, VOIP sensor, Accelerometer sensor, Artificial Intelligence. 1. INTRODUCTION Artificial Intelligence is a fundamental theory and development of computer systems able to perform tasks normally requiring human intelligence, and thinkingsuch as visual perception, speech recognition, face recognition, decision-making, and translation between languages. Nowadays the voice assistance what we use in our smart phones like "Ok Google” mainlyworks onAIandfurthermost the concepts such as Image recognition, Speech Recognition also works on Artificial Intelligence. [7] Now the next question that occurs in ourmindis howcanwe do this, what are the strategies used in this? And first of all how can we build machines with ‘commonsense’, the cognitive skills that every ordinary person takes it for granted? In this paper, we describe a common sense reasoning which is in the development stage through understanding some rudimentary flowcharts and some transition states of the flowchart. The system reasons the kinds of fundamental entities that show up in nearly all situations such as individuals, objects, occasions, objectives, plans, and slip-ups. We first describe the main features of our idea and then discuss its applicationandevaluationto an artificial life scenario. [4] 1.1 Existing This can be done with the help of Machine Learning where the computers are fed with the inputs, can begin, on its own, to see patterns and leads in these information,andfigure out how to perceive and arrange new information as they arrive into the patterns and rules that the computer has already created. After this these data are piled up and stored in a large database of a computer having its own rules and regulations making it more dependable [8]. 1.2 Proposed It our Idea as to how new types of future machines can be developed as shown below. Where we have briefed some ways how this concepts can be implemented to the real time application. Some of the paradigms of automatic common sensing machines are as follows. The first scenario is in our rooms where the moment we return from our work we will be sweaty or shivering of cold due to outside temperature and of courseoneofthese.There are some automatic AC’s which are hungonthewall whereit will detect our body’s humidity and set the AC according to our body’s temperature. Some thermal sensors are attached to that so in order to detect our body temperature. If in case our body temperature is low, then AC is set to the little more than the room temperature which would be suitable to his body temperature and opposite in the case when the body temperature is high. Secondly what we are working on right now is the common sense. Let’s try to induce common sense to Dustbin and make even that to have a sense too. 2. HOW DOES IT WORK? Well a lot of common sense is requiredthough, initiallythere are way too many dustbins in the world but our dustbin is a way too ahead. In the pas followed. For instance we have to bring our hand near it, the t days when we wanted to dump anything we used to go near the dustbins open it up and then we used to drop the waste there, but still as of now we still have tomove near the dustbin. But we induce common sense to it, then only if we move near the dustbin the dustbin should not open if in case if we have something in our hand it will sense it and opens .But there is an important thing here now, the dustbin thinks that whatever is in your hand as garbage so it blindly opens for everything there are some strategies to be
- 2. International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056 Volume: 04 Issue: 11 | Nov -2017 www.irjet.net p-ISSN: 2395-0072 © 2017, IRJET | Impact Factor value: 6.171 | ISO 9001:2008 Certified Journal | Page 557 point is what’s there in your hands do matter a lot. Whether good or bad things matters a lot. Choiceiswhetheryou want to throw it or not. Figure 1. Advanced automatic Dustbin A brief description about the diagram. As shown in the Figure 1. Well, it consists of a well equipped dustbin with embedded sensors to detect the input. And a lid on top, it’s quite obvious and depends on the user’s choice whether it is needed or not .In this case we use a lid just to make things simple. We don’t have to perform some extra things nevertheless just keep the hand above it ,There is no need of even touching it to be more precise its commonsense and it has a sensor.[3]. 2. 1 FLOWCHART AND DESIGN. The Flowchart Below depicts the following working of any machine which works on the basis of common sense. Here including start and stop there are additional of 7 states as follows The States are 1. Giving Inputs to the machines: This state involves various types of inputs given by the user to the machine be it the Hand gestures, through voice modulated commands and many more… 2. Checks for the Valid or Invalid one: This state involves evaluating the inputs whether is it a valid one or not. This further consists of. 2a.If at all the input is valid then it recognises the speech via speech or face recognition 2b.If in case the input is invalid one then it displays a suitable message stating that it is working on it. 3. Matches the input command and checks it in the stored database: This phase as the name suggests checks the command whether is it stored in the database and performs the operations according to it. 4. If the Command matches: then it performs two cases further. 4a. If matches: Then it performs the desired operations that meet the user requirements. 4b. If No: Then it returns to the case 2b and displays the suitable message. Most of the people think is it required. Whetherthecommon sense must be given to the machines or not or someareeven confused. Well the answer is now we are Presently in the midst of 2017 and as we can see slowly the things are becoming automotive such as the robots painting the car in the assembly lane, Drilling etc. Robots in the present have been a great asset because they canperformjobsthataretoo risky for humans. [6] Later on benefit robots are relied upon to outperform the mechanical Robot showcase. Numerous makers are beginning to investigate fields that are inadequate with regards to specialists. One zone of intrigue is creating automated aides for the elderly. You may likewise observe real employment of apply autonomy in law authorization and security. Future robots could likewise be as little as a PC chip and embedded into apparatuses, for example, vacuum cleaners. If this further continues then one day everything in this world would be automotive like automatic cooking machine the moment you return to home from work itwill senseyour hunger and prepare a dish for you, Ambient Light system is when u enter into the room automatically the lights will be turned on or off .This is how in further these machines will be dominating the world.
- 3. International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056 Volume: 04 Issue: 11 | Nov -2017 www.irjet.net p-ISSN: 2395-0072 © 2017, IRJET | Impact Factor value: 6.171 | ISO 9001:2008 Certified Journal | Page 558 2. 2 GRAPHICAL REPRESENTATION. As shown in the graph the machine is a lot more ahead than the human .It is so believed that when we teach a machine to perform some operations the better results can be obtained compared to that of a human being on the graph it takes a lead, and even the quality of the product developed by the machine would be a lot more better than human being. So overall teaching a machine gives a lot more sense that teaching a human .The reason why human beings are losing the race because they become lethargic, and often feel bored when doing the same work back and forth. Due to this the efficiency decreases, but on the other hand the machine never feels bored unless and until the work is finished it keeps on doing its work and surprisingly it never takes a break unless it has to do so. And the speed matters a lot because the machine works at very great speed when given the input compared to man. 3. ADVANTAGES It makes life simple by carrying all the tasks It is very user friendly and easy to use in terms of all aspects. Communicating with humans in a simpler way. 4. DIS-ADVANTAGES It is a bit cost effective. Understanding the working of the machine. Let me give an example “When you are conceived, you don't know anything.” But as the days go on and on we try to learn some rudimentary things through cognitive thinking, [5]. Actions which our parents do, and also through their conversation. In the same way the machine must be built first and there in the development stages through image processing it will try to identifytheimagesandtheirpurpose so when the user provides some input at the same time the machines tries to think which image is what and what must be done next as shown in Figure 3. So there will be something known as visual cortex which is one of the important Aspect in the machine which provides the ability to think. Where there are some preliminary steps or you can also it as an Algorithm. [2, 3] 1. Recognize the input 2. Process the input 3. React to the input 4. Perform operations according to the necessary Steps. 5. CONCLUSION The proposed system can be used to achieve simple and tedious tasks through normal human interaction. The commonsense to machine is growing to such to such an extent that even Face book; Microsoft is trying to implement it. As indicated by the current review Face book AI look into a mass chief Yann LeCun, jumps forward in neural systems administration and machine vision might one be able to day prompt programming with the presence of mind. Neural systems are simulated frameworks which imitate the structure of the human mind and by consolidating this with further developed machine vision which are approaches to pull information from symbolism for use in undertakings and basic leadership - LeCun says good judgment will be the outcome. For instance, on the off chance that you have a prevailing article in a picture, and enough information in protest classes, machines can perceive particular items like mutts, plants, or autos. But I promise this field has a promising future at great heights. REFERENCES [1]Davis E: ‘Representations of common sense knowledge’, San Mateo, CA, Morgan Kaufmann (1990). [2] Minsky, M.: The Society of Mind. Simon and Schuster, New York (1986) [3] Cambria, E., Hussain, A., Havasi, C., Eckl, C., Munro, J.: Towards Crowd ValidationoftheUKNational HealthService. In: ACM WebSci, Raleigh (2010) [4] Liu H., Singh, P.: Concept Net: a Practical Common Sense Reasoning Toolkit.BT Technology Journal 22(4), 211–226 (2004) [5]Speer,R.,Havasi,C.,Lieberman,H.:AnalogySpace:Reducing the Dimensionality of Common Sense Knowledge. In: AAAI (2008) [6] Havasi, C., Speer, R., Pustejovsky, J., Lieberman,H.:Digital Intuition: Applying Common Sense Using Dimensionality Reduction. IEEE Intelligent Systems 24(4), 24–35 (2009) [7]Singh, P.: The Open Mind Common Sense Project. KurzweilAI.net (2002). [8] McCarthy, J.: Programs with Common Sense. In: Teddington Conference on the Mechanization of Thought Processes (1959).