Decision support system for promotion of residential apartments in Chennai city using...

Decision support system for promotion of residential apartments in Chennai city using GIS

SHARE

S. Raghavendran
Pixel Infotek Pvt. Ltd.
6 DNR Layout, Palace Guttahalli
Bangalore-560 020, India
[email protected]

Over the last 350 years, Chennai has evolved from a group of fishing hamlets and villages into a vibrant metropolis. The locus of new residential developments is the south and the west, in an ever-growing semi-circle with the centre gradually moving southward. Major share of the housing supply is provided by the private sector. In the last two decades, development of apartments has become popular because middle-income group can afford only apartments and not independent houses. A new class of realtors known as apartment promoters have emerged, who construct the apartments and sell them. This has resulted in mushroom growth of apartments with a plethora of accompanying problems. With multiple housing finance options available from various lending agencies and interest rates at all time low, property buyers are sitting pretty today to get the best bargain in the market.

Need for the DSS
One of the biggest commitments anybody will make in one’s life, both financially and emotionally, is buying a house. Though it is interesting for a few, the experience of most of the buyers is fraught with frustration, exhaustion and poor satisfaction. A significant feature is that the location of a project plays a predominant role in the decision making process. Today buyers may get the right product at the right price but may not necessarily be at the desired location. This trend has discouraged many buyers to retreat from the market. The reason: non comprehensive database available about the apartment promoters and their projects, so that the buyers can choose from a variety of choices available to them. This being the case, no systematic effort has been made in our country, to collect data and to create a Decision Support System, to guide, control and facilitate the apartment promotion activities. The function of a DSS is to help the decision-makers (in this case the apartment buyers and the promoters) in taking quick decisions with easy access to the data with respect to which decisions are taken. This highlights the need for a DSS very much different from the conventional systems, in which data are scattered making it difficult for the decision-makers to have an access to reliable data and in taking quick decisions at the right time.

Objectives

  • To build a data base of residential apartments for sale within Chennai City
  • To build a database of sites for constructing residential apartments within Chennai City
  • To tailor the database created as a Decision Support System (DSS) for the following purposes:
    • To help the prospective residential apartment buyers in making their choice after analysing a variety of alternatives available to them.
    • To help the residential apartment promoters in selecting a site for constructing residential apartments.

Table 1: Average Weightage used in the Computation of Composite Site Index (CSI) on a 10-point scale

Sl. No. Factor Average weightage for each factor on a 10 point scale
1 Proximity to city centre 9
2 Land Cost 5
3 Availability of Ground Water 9
4 Availability of Metro Water 8
5 Availability of Sewerage system on the abutting road 8
6 Site to be free from Inundation 4
7 Proximity to Educational Institution 6
8 Proximity to Railway Station 5
9 Proximity to Bus Terminal 6
10 Proximity to Hospital 6
Source: From the analysis of the questionnaire survey with the promoters

Methodology
In order to achieve the objectives set, the following methodology was adopted:

  • Defining and understanding the current problems of the buyers and the promoters
  • Delineating the study area or selection of study area
  • Data Collection: Questionnaire survey to be conducted with the promoters to identify the factors dictating site selection for constructing residential apartments and the weightages given to those factors by promoters.
  • Information regarding residential apartments for sale and sites for constructing residential apartments to be collected
  • Analysing and ranking the apartments for sale and the sites for constructing residential apartments for sale included in the database for calculating the Composite Apartment Index (CAI) and Composite Site Index (CSI) respectively
  • Creation of base map of the study area using GIS and creation of database from the information collected
  • Integrating the spatial and non-spatial database created and building the DSS
  • Web launching of the DSS
  • Sample queries and validation of results.

Study Area
The prime justification for selecting Chennai City as the study area is that more number of residential apartment projects are promoted only within Chennai City, while layouts or vacant plots are more popular in the suburban areas. Thanks to the multiplicity of housing finance institutions offering finance at competitive rates, the MIG (Middle Income Group) and LIG (Lower Income Group) working in the government, banks and public sector undertakings can now afford to purchase an apartment within the city and see their long cherished dream come true. In fact this has increased the demand for LIG and MIG housing

Data Collection
Collection of relevant and accurate data is very essential for building a good database, which is supposed to be the backbone of any DSS. The required primary data has been collected through field survey and the secondary data through various other sources.

Decision support system for promotion of residential apartments in Chennai city using GIS

Primary data
Various primary data collected through field survey are:

Information regarding residential apartments for sale and sites for constructing residential apartments within Chennai City. This information was collected from a variety of sources such as Newspapers, Internet websites of the promoters, Telephonic interview with the sellers, Property Fair ’99 exhibition etc. The following are the information that were collected regarding residential apartments for sale within Chennai City:

  • Name of the Agency promoting the Apartment, Contact Address, Web Site Address, Email Address, etc.
  • Location and address of the Apartment for Sale
  • Apartment Cost per Sq.ft
  • Key Plan, Site Plan, Floor Plan and Photo of the Apartment
  • Additional Information about the Apartment and the Agency such as the Special Features of the Apartment, Profile of the Agency, etc.

The information that was collected from the sellers of sites suitable for constructing residential apartments are:

  • Name of the Seller
  • Contact Address
  • Site Area in Sqm
  • Cost per Sqm
  • Site Location and Address
  • Brief Description about the Property
  • Willingness of the Seller for Joint Venture
  • Ground Water Availability (Depth in Feet ) and Drinking Suitability
  • Whether the Property is with/ without Building, etc.

Weightage given for various factors by apartment promoters when selecting a site for constructing residential apartments through questionnaire survey with the promoters. The following are the factors for which the promoters were asked to assign weightage on a 10-point scale:

  • Proximity of the Site to City Center
  • Cost per Sqm of the Site
  • Availability of Ground Water (Potability)
  • Availability of Metro Water
  • Availability of Sewerage System on the Abutting Road
  • Site to be Free from Inundation
  • Proximity of the Site to Educational Institution, Railway Station, Bus Terminus, Hospital etc

Secondary data
Various secondary data that were collected include Trend in Residential apartment prices (Rs/Sqft) for the past five years (1995-99) and Residential land prices (1 ground price Rs in lakhs) (1999-2000), at selected locations within Chennai City from the journal titled “A Guide to Chennai Real Estate 1999”.

Table 2: Procedure for Ranking Sites for Constructing Residential Apartments

Sl. No. Factor Criteria for assigning ranks Procedure for assigning rank
1 Proximity to city centre Distance in Kms Rank 1 for the site closest to city centre
2 Cost per M1 Rs/M1 Rank 1 for the site which is having the least cost per M1
3 Availability of Ground Water Depth in Meters and Potability Rank 1 for the site where water is available at a shallow depth and also potable
4 Availability of Metro Water Frequency of supply Every day = Rank 1
Alternate day = Rank 2
Others = Rank 3
Not available = Rank 4
5 Availability of Sewerage system on the abutting road Frequency of Blockage Rare = Rank 1
Frequent = Rank 2
Very Frequent = Rank 3
No sewage system available = Rank 4
6 Inundation in the area Frequency of Indundation No Inundation = Rank 1
Rare = Rank 2
Frequent = Rank 3
Very frequent = Rank 4
7 Proximity to Educational Institution Distance in Kilometers Rank 1 for the site having an educational institution very close to it
8 Proximity to Railway Station Distance in Kilometers Rank 1 for the site having a railway station very close to it
9 Proximity to Bus Terminal Distance in Kilometers Rank 1 for the site having a bus terminal very close to it
10 Proximity to Hospital Distance in Kilometers Rank 1 for the site having a hospital very close to it

Sampling for data collection
It needs no mention that the real estate sector in India is an unorganised one and it is a herculean task to collect information regarding the total number of apartment promoters operating in an area. This being the situation the scope of defining the sample space by using Stratified Sampling/Cluster Sampling technique gets defeated as these techniques require prior information about the promoters such as the number of years the promoter has been in this field, number of apartments promoted so far etc. to stratify the population. So the population has been defined by exhausting the list of promoters in the Yellow Pages of the Chennai Telephone Directory, Members of FAIRPRO (Foundation for Fair Practices in Property Development), List of Promoters who attended the Property Fair’99 exhibition held at Chennai, the list of promoters having website addresses. All these were merged and overlapping addresses eliminated and the final population arrived at. Random sampling has been done to select the sample from the population.

It is worth mentioning here that, eventhough attempts were made to collect information regarding the approval status of the sites and apartments for sale (i.e. the Development Control Rules (DCR) compliance), all efforts went in vain, as such information is possible to collect only for an authority like CMDA (Chennai Metropolitan Development Authority).

Data Analysis
“Method of Weights” is a very common method in urban analysis for calculating Composite Site Index (CSI), to identify the best site among the available alternative sites based on a set of factors, which are supposed to have an influence in the process of site selection. The method involves fixing of weightages to various factors involved in the calculation of CSI by Expert Opinion Survey/Delphi Technique/Questionnaire Survey. Then on the basis of the factors considered for calculating the CSI, the sites are ranked in comparison to one another. For a given factor the top most rank is assigned for the best site and the bottom most rank for the worst site. This is carried out for all the factors considered, for all the sites. Then for each individual site, the rank assigned for that site for a particular factor (say for e.g. Proximity to City Centre) is multiplied by the corresponding weightage for that factor as obtained from the Expert Opinion Survey/Delphi Technique/Questionnaire Survey, etc. The summation of these products for a particular site is nothing but the Composite Site Index (CSI) value for that site, based on the set of factors taken into consideration. Composite Apartment Index (CAI) and Composite Site Index (CSI)
CAI/CSI is a combination of 10 factors on the basis of which it has been arrived at. Any buyer/promoter who does not have any special preferences over any one aspect, when making a decision to buy an apartment/select a site for constructing residential apartments, but is satisfied with an apartment/site which in comparison with the other apartments/sites listed in the database, is at close proximity to facilities like educational institutions, hospitals, railway stations, bus terminus; close to city centre, etc. The formula for calculating the Composite Apartment Index for an individual apartment for sale is as follows:

Where, i = Number of apartments included in the database (in this case it is 30 )

The formula for calculating the Composite Site Index for an individual site is as follows:

Where, i = Number of sites included in the database(in this case it is 15 )

Ranking sites for constructing and sale of residential apartments

In order to work out the rank for various factors used in the calculation of CSI, for each individual site/apartment included in the database, the formula indicated earlier has been used. The weightages for the factors (Table 1) have been arrived from the analysis of the questionnaire survey conducted with the promoters. Taking into consideration the factors, same as those used for the computation of CSI, the 15 sites included in the database of sites for constructing residential apartments for sale within Chennai City have been ranked to arrive at the CSI for each site. The factors considered and the technique adopted for assigning ranks are as shown in Table 2. The ranking obtained in this manner for each factor for each site was then multiplied by the corresponding weightage for that factor to arrive at the CSI for that site. In ranking the residential apartments for sale, the procedure similar to that adopted for ranking the sites for constructing residential apartments has been used, with a small difference in the criteria for ranking the apartments in terms of availability of ground water. The criteria has been evolved taking into consideration the availability of ground water for purposes like all purposes, drinking and cooking, ablution purposes only and finally scarce availability of ground water. The following ranks have been assigned for availability of ground water different purposes:
v All purposes especially Drinking and Cooking (Very good condition) = 1

Drinking and Cooking only (Good condition) = 2
Ablution purposes only (Satisfactory condition) = 3

Scarce availability of ground water (Scarce condition) = 4

Building the DSS
The DSS has been built taking into due consideration the User, the System, the tool selected (GIS), and finally the problem to be addressed. The components of this DSS can be divided into two types, viz. Spatial and Non-Spatial.

The spatial component involves base map creation, where the graphical entities are represented spatially. The non-spatial component involves designing the database where the relevant information about the graphical entities is to be stored and retrieved for analysis.

Base map creation for the study area
A 1:25,000 scale TTK map of Chennai City was scanned and imported in the raster format for creating the base map of the study area, through onscreen digitising. MicroStation Descartes of Bentley Systems. has been used for semi-automatic R2V conversion. Using GeoDefiner, the coordinate and projection systems have been defined for the vectorised base map. The point features that have been included in the base map, have been geopositioned by keying in the latitude and longitude of the respective features. This is very much essential for accurately locating the position of new apartment projects on the map. The partial list of various features that have been included in the base map are: Roads, Railways, Water Bodies, RailwayStations, Bus Terminus, Hospitals, Educational Institutions, Greenery, various apartment projects, sites for new apartment projects etc. (Fig. 1)

Database design
Database is a subset of data elements taken from a databank and organised for storage and manipulation in a computer system. Database is the fulcrum of any Decision Support System. With the advancements in information technology, access to large and remote databases, hitherto available in the form of a CD-ROM, have now switched over to GIS maps. When data from different sources can be displayed in many different ways at the press of a few keys, then it is possible to gain insights and detect patterns which could not easily be achieved by other means. This is possible only with a tool like GIS.

In this project the database has been built using MS-Access. Non-Spatial database is nothing but the non-graphic information about a feature, stored in a table and linked to the feature. The residential apartments for sale represented graphically on the base map, contains the following attribute information:

  • Name, Contact Address of the Agency promoting the Apartment
  • Apartment Location and Address
  • Apartment Area in Sqft
  • Apartment Cost per Sqft
  • Key Plan, Site Plan, Floor Plan, and Photo of the Apartment
  • Distance to City Center
  • Name of the Nearest Educational Institution
  • Distance to Nearest Educational Institution
  • Composite Apartment Index (CAI) for that Apartment
  • Rank based on CAI for that Apartment
  • DCR Compliance of the Apartment, etc.

Similarly the following is the partial list of attribute information are available for the sites for constructing residential apartments for sale included in the database created:

  • Name of the Seller and Address
  • Site Area in Sqm
  • Cost per Sqm
  • Site Location and Address
  • Willingness of the Seller for Joint Venture
  • Name of the Nearest Educational Institution
  • Distance to Nearest Educational Institution
  • Ground Water Availability (Depth in Meters (Feet) )
  • Drinking Suitability of Ground Water (Yes / No)
  • Whether the property is with / without building
  • Land Price (One Ground Price) at selected locations within Chennai City (1999-2000)
  • Composite Site Index (CSI) for that Site
  • Ranking based on CSI for that Site
  • DCR Compliance of the Site.

Additional information about an apartment has been created using MS – POWER POINT so that a buyer after selecting the apartment can go through this video show to get additional information such as materials used for various works, special features, other projects of the promoter of this apartment etc. The key plan, site plan, floor plan and apartment photo have been stored as image files in a separate directory and hyperlink given to the respective files. The trend in residential property prices has been shown in a graphical form by using Bar Graphs in MS-POWER POINT and Hyperlink has been given to the respective file under the attribute value field. Database Integration
The process of establishing a linkage between a spatial feature located on a map, to its respective non-spatial or attribute information is termed as database integration. For this purpose MicroStation/ J Geographics (Ver. 7.1) has been used for project creation and database integration.

Web launching of the application
Using Model Server Discovery, the web component of MicroStation/J Geographics the application has been customised so that the end users can make queries on the DSS on the web using Internet Explorer or Navigator or other popular web browsers, anytime, anywhere. By using SVF (Simple Vector Format) type of image publishing on the web, it is possible for the end user to zoom in/out, turn on/off the layers etc. dynamically.

Sample Queries and Validation of results
In the DSS built it is possible to make three types of queries. They are:

  • Attribute Filter Queries
  • Spatial Filter Queries and
  • Attribute and Spatial Filter Queries (Combination of the above two queries).

A query in GIS terminology is a request for information from the database built. In an attribute-filter query, it is possible to identify the features desired by defining an attribute filter. This can be accomplished using the visual SQL query builder. On the other hand Spatial queries can be made by using the generate zone tool available in MicroStation/J Geographics. A combination of the above two types of queries can also be made.
A typical attribute filter query is given below
NAME_OF_THE_AGENCY LIKE ‘A’ AND APARTMENT_LOCATION = ‘TNAGAR’ AND AREA_IN_SQM (SQFT) > 500OR COST_PER_SQM (SQFT)