Wednesday, August 01, 2018

Kudos to Kerala on Efficient Handling of the Nipah Virus Outbreak


Visit Nipah virus: Anatomy of an outbreak a report in the Hindu newspaper.

Kozhikode: Kozhikode and Malappuram districts, where 17 people died of Nipah in May, were declared free of the virus by the Kerala government in July 2018. "A decision to declare both districts as Nipah-virus-free was taken by the government as no fresh cases have been reported after June 1," health minister K.K. Shylaja said. 
She was speaking at an event to felicitate medical teams, support staff and volunteers who fought against the spread of the virus.
Sajeesh, husband of 28-year-old nurse Lini Puthuserry who died after contracting the virus while attending to patients at Perumbra Taluk Hospital, was among those honoured.
The Collectors of Kozhikode and Malapurram districts, U.V. Jose and Amit Kumar Meena, and Manipal Virology Institute director G. Arunkumar were also honoured. Dr A.S. Anoop Kumar of the Kozhikode-based Baby Memorial Hospital was presented with a special award for his efforts to identify and prevent the virus outbreak.
The government had on June 11 lifted the travel advisory it had issued in the wake of the virus and said it was safe to travel to any part of the state.
It is heartening to see how they have managed the crisis efficiently. It is frightening to think of what could have happened if the outbreak had happened in a place where healthcare is not as efficient as it seems to have been in Kerala.
The speed with which Kozhikode’s Baby Memorial Hospital seems to have ruled out alternative hypotheses and suspected Nipah infection is impressive. The availability of a very competent critical care doctor and an equally competent neurologist there to handle the first set of Nipah victims speaks highly of the state of health care in Kerala. 
Credit for the second impressive feat belongs to a private institution in Karnataka – which has been operating the Manipal Centre for Virus Research (MCVR) - and to Dr Arunkumar, Prof and Head of the Department at MCVR. In 2017 MCVR had sent a team to the US Centre for Disease control. One of the actions taken was training some members of the visiting team for testing samples to identify the presence or absence of the Nipah infection. This test can be currently done only at two places in India: National Institute of Virology, Pune and MCVR. 
The Kerala hospital sent samples to MCVR, 300 KM away, for quick identification of the virus involved. MCVR under the Manipal University has also been supported by the Govt of India. This public-private partnership in the critical area of public health shows what can be done in India when dedicated leadership is available.

India as well as the World Health Organization have done the right thing by honouring Lini Puthussery a dedicated healthcare professional who died after taking care of Nipah victims. We need to recognize that the professionals and other staff who work in public health are some of the most valuable people in India. It is necessary to institute annual awards to honour these committed workers.  We must also identify hospitals that maintain readiness to deal with public health emergencies.  

Srinivasan Ramani  


References

1) Nipah virus: Anatomy of an outbreak, Priyanka Pulla, JUNE 02,   
     
2018 00:15 IST, UPDATED: JUNE 02, 2018 12:38 IST

2) WHO pays tribute to Kerala nurse who died of Nipah, PTI,
      Jun 05, 2018, 05.12 PM IST



                                                                      

Saturday, May 19, 2018


The Hanging Rules
There was a time when algorithms were considered more important if they did not depend on the nature of data they processed. Take sort algorithms, for instance. They work just the same whether you sort the weights of a set of pigs, or the age of a set of galaxies. I have no quarrel with such algorithms; in some sense they deal with the highest level of abstractions such as numbers.

However, nowadays the challenges are more in dealing with large volumes of data from different application domains. Knowledge related to the domain are basic to success with most of our endeavours. Computations of this type do not necessarily lead to unique solutions to a given problem. However, the solutions they provide could have great value. This value arises from the importance of problems they help us solve.

Why this build up for what is to follow? It is because I wish to discuss a very important problem. It is the problem of shaping pre-poll alliances among political parties. It is clear that the Indian election to be held in 2019 will critically depend upon pre-poll alliances. The recent Karnataka election has high-lighted the significance of alliances. Commentators have pointed out that had there been a pre-poll alliance, the outcome would have been quite different. Pre-poll tie-ups will be important for all parties: those with a high vote share as well as for the smaller parties. When vote shares are added together, the winner may not get the aithmetic sum of the two vote shares. When two parties with vote shares (expressed as percentages of votes cast) A and B have an alliance, the winner does not get the vote share A + B but may get something like A + B/2. The effect on the election results could however be very dramatic. Pre-poll alliances have demonstrated their effectiveness in the past in this regard. Increasing your vote share by B/2 does not just increase your chances of winning by 50%. It may increase it 500%.  

Given this logic, why do parties hesitate to enter into pre-poll alliances? Partly because reliable predictions of possible gains are not available. Secondly, there are no clear procedures which can be used by the parties concerned to decide on how they should share constituencies between them. How many constituencies would each party get? Which constituencies would be assigned to which partner? These are complex questions, even when only two parties participate in a pre-poll alliance. Multi-party computations are more complex, but the generalization form a 2-party to a multi-party situation is not very difficult.

The major difficulties in persuading political parties to consider this problem analyticaly is that solution should be fair to the potential allies and must be seen to be clearly unbiased. The solution should also be easy to communicate and understand. The solution cannot be thrust upon the parties involved and must leave some room for tweaking as per their intuition and perceptions. I will offer a solution here to the two-party problem and leave it to the reader to generalize it to the multi-party situation. I don’t mean to suggest a solution that should be adopted blindly; I only wish to have many people think out their own versions of the solution. Secondly, I want to keep the solution very simple. You can make it more accurate as a predictor of the gains of a pre-poll alliance by making it more complicated, but very few political leaders are computer scientists. It is difficult to persuade them to adopt what sounds like mumbo-jumbo to them!

Lastly, why the funny name? It comes from the saying “If you don’t hang together, you will hang separately”! So, this is a solution that shows how some politicians could hang together!

The solution: 
  1. Assume that the vote share received in each constituency in the last election is the best estimate of the vote share that a party would get in the forthcoming election.  
  2. If the estimated vote share A of one of the two parties is greater than that of the other B, assign that constituency to the party with the higher estimated vote share.
  3. Compute the total A + B/2 where B is the vote share estimated for the other partner. Let C be the highest vote share estimated for the strongest party outside the alliance. If (A+B/2) > C, mark that constituency as a strong seat for the proposed alliance. The underlying assumption is that an alliance does not transfer all votes of from one partnering party to the other. For simplicity, we assume that 50% of the votes can be "transferred". (carry out the above three steps for each constitutency).
  4. Make adjustments to the extent of 20% of the seats by negotiation, provided this does not disturb the assignment of any strong constituency. However, the number of seats assigned to partners should be in proportion to the total vote shares each has earned.
  5. The above rules are not written in stone. But if the two partners cannot agree to some reasonable modified version of the same, they can hang separately!

          Srinivasan Ramani       

Saturday, March 31, 2018

A Rating for People's Representatives!

Many of us are now used to rating goods and services. The dosa you had in a small restaurant seemed to be the best you have eaten in some time? You write a few lines of review for a site that rates restaurants and give it 4.5 or 5 stars! You thought the driver of the Uber or Ola cab you took to the station could have been more helpful. Ok, give him only three stars!
What about you CM, or your MLA, or your councillor? Why not rate him on a number of dimensions? Voting for/against them once in five years is not good enough. We ought to share our opinions about their deeds and failures almost every month.
The way a leader is to be judged is no doubt very important, but as this is a blog in the field of science and technology, our first priority would be technical issues in designing a rating website. I would suggest the following design objectives:
Credibility: it should be difficult to fool the website and manipulate the data going in
Authentication: The site should ideally work on “one citizen – one vote” principle
Privacy: People who provide ratings may not want some shady company or hackers, to sell a record of their ratings. A good design involves careful consideration of privacy on a site which insists on authentication. One simple design would be to have ratings come in only through SMS messages. No prior registration is required. We can ensure that a cell phone can vote on an issue only once. It can give a rating only to the persons/offices in the state from where it sends a message. I will assume that most users do not care too much about hackers finding out how one’s cell phone voted. After all, in most countries, we are not so scared of the government as to keep our political opinions a secret.       
Simplicity: The system should be based on a good trade off – not too complex to drive users away and not so simple that it can be fooled easily
Accessibility: The system should be accessible from a smart phone. This will deny access to a significant fraction of the Indian population, but it would enable you to build a working system easily. The trade off in this dimension could be examined and different choices can be explored by different groups. What is the value of a smart phone when the votes are cast by SMS? The user can use a smart phone to read the web pages which will help the users to select the issue/person they wish to rate. If the web pages support it, they could answer multiple questions giving answers in the form of a numeral in the range 1-5. The system could summarize answers to six questions, for example, as 325-45 meaning that the six questions were answered as follows, giving ratings on a scale of 5, where 5 is excellent and 1 is very poor:

Question 1: Answer 3
Question 2: Answer 2
Question 3: Answer 5
Question 4: Skipped
Question 5: Answer 4
Question 6: Answer 5

The website would also give a unique ID to the person or office rated, for example SRM for Srinivasan Ramani. The user has to send an SMS to a given number reading, for example as
   SRM 325-45
Why send an SMS, and not submit the answers on the Webpage itself? The security mechanism provided by the cellular network identifies the sender’s cell phone number reliably. In my opinion it is reliable enough for the application and is simple enough to implement. It is also worth noting that the cellular network sends with the SMS the sender’s phone number, the time of sending, and the location from which the message is sent. A computer receiving the messages can use all this information to process them suitably.
The answers I have given to some of the design questions above are not binding on you. You may make your choices. A system like this makes sense only when it operates with a large number of users. So, you need to plan for scalability and reliability. The systems should work round the clock, round the year. Lastly, you need to decide on what your business model would be? Who would pay the bills after the system scales up?
Now, we can turn to the nature of comments and ratings we would like to collect. I have some suggestions about what questions we can ask about representatives and service providing offices but again, in your project, you are free to make your own choices. Please treat my questions as examples what you could ask.
Questions about a people's representative
·        Is he a visionary leader?  (Here and in the following, I merely use “he” to refer to the representative. It can obviously be a “she”).
·        Has he been effective in getting a few major problems solved?
·        Is he known for his integrity?
·        Does he have cronies or sons who are not exactly assets to society?
·        Does he communicate his ideas well?
·        Does he do this frequently?
·        Has he kept his electoral promises?
·        Are you likely to vote for him next time?

Questions about specific issues
·        Should your state have prohibition?
·        Should we ban children below ten being given rides on two-wheelers on city roads or highways?
Questions about specific offices that provide a service
·        Is a given bank branch helpful and efficient?
·        Is a given college a good place to send students to?
·        Is there corruption in a given hospital?
As, I see it, each of these questions is a like raw nerve. The pain-signals of our lives travel through these nerves, but do they bother the person who is responsible for the pain? Usually no! The nerve does not run far enough to touch him.  
Making the ratings public, and highlighting them during election time, are important to make the system sit up and take notice. 
Getting people to rate their representatives and their governments is not exactly new. You can see what other websites offer in this area. Visit
There is a lot of room for innovation, but the basic idea is simple: public feedback makes the high and mighty accountable. App-based taxicab services are responsive to people's problems. We should demand that people's representatives and government services be equally responsive.

Srinivasan Ramani

Thursday, March 15, 2018

Use of dictation software to teach speech and hearing-impaired


There are millions of deaf children in India. Visit Varshney S. Deafness in India. Indian J Otol, 2016; 22:73-6. http://www.indianjotol.org/text.asp?2016/22/2/73/182281
It is generally believed that those who are hearing-impaired need not be speech impaired. They  cannot hear themselves speak and cannot hear spoken responses from others they are talking to; so, they fail to get feedback on their own speech. This absence of feedback is considered the major factor in the hearing-impaired usually not learning to speak. This raises an interesting question: can we encourage students with this problem to speak into a cell phone running a dictation app to get useful feedback on how their speech is perceived by others? A cellphone with a suitable app is not automatically going to solve the problem. A capable teacher, at home or at school, would usually be important to teach the hearing-impaired to use the tool. Further, the learner should have a working knowledge of English.  The ability to learn more will depend upon the level of language skills at that time. So, we must expect that the side-by-side learning of English and use of this tool will be a process that will take months or years of sustained work. Trained teachers will of course speed up this learning a lot. Visit https://www.audiologyonline.com/articles/auditory-oral-education-teaching-deaf-1248
Dictation software of pretty good quality is available on devices such as iPhones and Android phones. I believe you can get cell phones costing less than Rs 4000 (less than USD 70), that can run dictation software. It would be valuable to choose a rugged phone which can stand a few drops, as a broken cell phone can often kill the user’s hope and enthusiasm. A display with a diagonal dimension of  4 inches or more would be useful for this application.
We have to recognise two advances: improvements in the performance of dictation software and wide availability of cell phones to run such software. Also significant is the availability of apps such as WhatsApp that take spoken text messages and send them off to other cell phone users. We empower the hearing-impaired when we give them the ability to send easily created text messages through their cell phones. This enables them to communicate with friends and family irrespective of distance, and without struggling to make a voice call.
The technology discussed above works equally well in enabling people to communicate to the hearing-impaired. The speaker can dictate into any cell phone what they want to say and show the text to the “listener”.


It is worth noting that the cost of the hardware involved is significantly less than that of a good-quality hearing aid, or the cost of a cochlear implant. This is not to argue against hearing aids which have their own value. We should visualise hearing-impaired students using hearing aids as well as cell phones equipped with speech recognition apps. They will complement each other, and together they would enable the users to cope with a whole variety of situations. It is worth noting that hearing aids do not completely solve the problem of poor hearing. Very often hearing-impaired students use affordable hearing aids which do not provide the necessary high quality of user experience. Complementing them with speech recognition apps on cell phones is a good way to to reduce the problems faced by the students, and to speed up their learning.


Can we use an Indian language in this context? I have only focused on English in this article, but Google Docs is able to cope with 40 different languages (as of March 2018), irrespective of your using it on a computer or on (an Android)  cell phone. The one Indian language included in this list of 40 is Hindi. However, I have not tried voice typing in Hindi with Google Docs. Any user interested in this should expect to spend some time learning relevant information and skills. Postings on the Web say that your Google Doc app should be up-to-date and that you should go to settings, then languages, and add Hindi. I am sure that good dictation software will become available in all major Indian languages over the years.  

Lastly, the most wide-spread use of dictation software in India would be among users with normal hearing who wish to improve the way they speak English.


Srinivasan Ramani

Tuesday, March 13, 2018

The search drone problem


Imagine that a search drone is on the look-out for a submarine that is careful to avoid detection. The controllers of the drone have been given a convex polygonal area on the surface of the sea and told that the drone should search the area inside that polygon. Assume that the drone flies at 150 KM per hour and the polygonal space covers roughly an area of 60,000 Sq. KM. Further assume that the drone flies at a height of 1 KM above the sea and its sensors can detect the submarine if it is inside a circle, 1 KM diameter, with its centre directly beneath the drone.  
Your task is to program a flight pattern for the drone. You can assume the availability of a function in(X, Y) in the control computer, that returns 1 if the drone’s current position (X, Y) is inside the polygon and 0 otherwise.
It is desirable that the drone’s path be relatively unpredictable, and that all places inside the polygon have equal chances of being visited to the extent possible.
Can you write a program that would plot the scan pattern showing how it would fly the drone over the given area? Use any graphical display software to provide a dynamic display.
Assume that the maximum underwater speed of the submarine is 10 KM per hour when it is trying to evade detection. What would be a good strategy for the submarine to avoid detection under these circumstances? Assume that the submarine gets to sense where the drone is, if it gets within 4 KM of horizontal distance. Try programming this strategy and show the path of the submarine on the display in a different color. 

Srinivasan Ramani