Summer Research Fellowship Programme of India's Science Academies 2017
Table of contents
Introduction 3
Related Work 4
2. Data collection 5
Web Scraping 7
3. Mobile Application 8
Offline Maps 12
4. Web Portal 13
Path planning 14
Path Rendering 15
LP Formulation-I 16
LP Formulation - II 17
Testing and Analysis of User feedback 19
5. GPS logging 20
6. Future Scope 21
7. Conclusion 21
8. Tools, Libraries and Resources
9. References 22
Tour Companion
There is a plethora of information associated with every tourist spot, however this
knowledge is not conveniently available to tourists at the attraction itself. The aim is to enhance
the experience of tourism by enabling tourists to effortlessly access this information at the right
place and time. There are many resources on the internet that make this information available
but there is one major problem with all of them, they all rely on the user to know what to search
for, which more often than not users do not know. The solutions proposed and developed
eliminate this requirement, by making relevant information available without the need for user
intervention. A mobile application and web portal was developed which makes the information
available and techniques like Natural Language Processing, Linear Programming and Video
Summarization using Computer Vision are used in the organizing of this information. The most
essential element of this enterprise is not to simplify the process of accessing information but
actually the collection and organization of this information. Another problem addressed is the
lack of reliable connectivity at remote tourist attractions which makes it impossible to use
modern conventional sources of information.
For on-site assistance a mobile application was developed, the mobile app developed is
completely offline and relies only on GPS connectivity for its services. As a proof of concept the
app was developed and tested for the Nehru Zoological Park, Hyderabad. However it was
ensured during development that the app is generic and can be used at any tourist spot subject
to availability of organized data.
For off-site assistance a web portal was developed which enables tourists to efficiently
plan their visit to tourist destinations. The portal also has a virtual tour feature which lets tourists
experience the destination from the comfort of their homes before actually visiting the spot. The
virtual tour is dynamically generated based on the constraints input by the user. The web portal
is also generic and can be used for any tourist spot by simply adding its database in the
prescribed format.
To improve the experience over time, the mobile application has a data collection feature
which the users can choose to turn off. This data collected is analyzed using data analytics and
visualization tools to improve the on-site and off-site services.
This report summarizes the data collection, development and testing phase of the mobile
application, the on site tour companion and the web portal, which serves as the off-site tour
companion. With these solutions developed we hope to enhance the experience of tourism by
making the tourists more informed of the places they visit.
Related Work
In the tourism app/website space there are numerous services to assist with travel
bookings, hotel bookings, translation, dining, transactions, popular opinion ,etc. but nothing that
gives tourists hyperlocal information about the attractions they visit.
Field Trip an app by Niantic attempts this and gives general description of places to visit closeby
but fails at giving information about the various points of interest at the tourist attraction. Field
trip also relies heavily on internet connectivity. There are many community driven travel blogs
and websites like TripAdvisor, Tourism Review and Tourism sites of local authorities which give
a good idea about the places to visit along with people’s experiences, but these again require
the user to look for information and often give mixed reviews about everything.
A lot of research has gone into the advent of the smartphone revolutionizing the tourism industry
and it has succeeded the commercial domain of tourism but has been limited in making the
tourist more informed about the places they visit.
The Tour Companion App accomplishes what existing apps have not that is, providing
hyperlocal information about the points of interest at the tourist spot without any user
intervention, another important feature is that the app does not rely on internet connectivity
which is not always reliable at all tourist spots. The limitation of this app is the lack of availability
of data which must be collected manually, this is a cumbersome and time consuming process. If
this process of data collection is simplified the Tour Companion App along with the tour
companion web portal can truly enhance the overall experience of tourism.
2. Data collection
The most vital part of the application is the database which has all the gps location of the
animals in the zoo with the description.So the first step was to collect data. For this the Nehru
Zoological park was visited multiple times. During these visits the data collection methodology
was finalized, this methodology is such that the data can be collected by simply visiting the spot
as a regular visitor with basic equipment that all tourists carry.
For the zoo park it was observed that GPS accuracy was around 1-3m for some animals and
10m for some animals due to bad connectivity. Description boards were damaged or faded and
not present at all for some animals, the number of languages these descriptions were present in
was also limited and for obvious reasons. These are the problems that one might encounter
while visiting any tourist destination around the world.For a short video of the point of interest,
animals housed at the zoo in this case it was noticed that a professional camera might be
required for a good quality video. This will be the case at any tourist spot, some points of
interests can be observed only from a distance. Videos for these points of interest can easily be
found from online resources without violating any copyright laws.This was done for dangerous