Topic Analysis of Cloud Legazpi Feedback Portal Open Data USAPANG CITY HALL (Qualitative Responses)

By: Lany L. Maceda, University of the Cordilleras, Doctor in Information Technology

From the qualitative responses given by the citizenry or clientele of the Local Government Unit (LGU) of Legazpi City, several Runs in topic analysis was made until a satisfactory composition was achieved.  In the end, topic labels with the numbers of topic models were manually labelled. For ambiguous words, I reevaluated both the raw and cleaned data and analyzed how they were used to determine appropriate labeling.

The data analytics was made from the (1) Remarks and (2) Messages posted by the citizenry on the Cloud Legazpi Feedback portal. The Messages given were classified according to its response type such as (a) LGU, (b) Public, and (c) Outside.

1. REMARKS

Raw Data = 134

After Cleaning = 83

• Cleaning includes removing redundancy, URL/Links, mentions (@user1), and hashtags (#) from data.

Test Run was set to:

Topics = 5

No. of Iterations = 5

No. of Words = 5

Optimization Interval = 5

Topic Labels	Topic Models	Topic Distribution
0 Employee attitude	empleyado, approachable, woman, serbisyo, servant	20%
1 Service Delivery	serbisyo, requirements, asikaso, pagkakataon, transaction	20%
2 Special Lane	documents, pila, pregnant, pupuntahan, gagawin	20%
3 Office Environment/ Community Assistance	office, matulungan, nabibigyan, maingay, time	20%
4 Clients/Customer’s Relation	maayos, lahat, medyo, staff, masungit	20%

Details:

Topics = 5

Topic Bits = 3

Topic Mask = 111

Maximum Tokens = 24

Total Tokens = 150

Time: 30 seconds each topic model

              Sample Remarks Posted on each Topic Label:

Topic 0: Employee Attitude

• “they are so approachable  and ready to be of help”
• “fast service and approachable staff :)”

Topic 1: Service Delivery

• “mabilis ang transaction”
• “nag level up na ang lgu legazpi city”
• “sapat lang ang serbisyo”

Topic 2: Special Lane

• “magandaang  kanilang serbesyo s pag bigay ng mga documento”

Topic 3: Office environment/community assistance

• “kahit mahaba ang pila at medyo maingay lahat naman ay nabibigyan ng pagkakataon na matulungan“

Topic 4: Client/Customer Relations

• “Ayos naman, kaso murusduton tabi he he he”
• “NGUMITI NAMAN”
• “Sobrang tagal magprocess. Masungit pa.”

Visualization of the “Remarks” Topic Models:

The remarks were given by the citizenry to the following offices of LGU Legazpi:

1.Business Permit and Licensing Office:

• Release of Business Permit

2. City Accountant Office:

• Issuance of Disbursement vouchers for claimstaken up as Accts. Payables
• Issuance
• vouchers/payrolls

3. City Administrator:

• Assistance to clients seeking queries advisescomplaints and other concerns
• For signature of Checks/ documents

4. City Assessor’s Office

• TRANSFER OF OWNERSHIP OWNERSHIP SECURING ASSESSMENTOF BUILDING MACHINERY AND OTHER STRUCTURES

5. City Budget Office

• Clients submission of payrolls vouchers PurchaseRequest (PR) etc. accompanied by Obligation Request (ObR) Form for funds availability

6. City Engineering Office

• Building Permit

7. City Health Office:

• Preparation Of Medical Certificate
• General Medical Services Nursing Division
• Pre And Post Natal Services
• Dental Services

8. City Human Resource and Management Office

• Issuance of Job Order Appointments & Contract of Services
• Recruitment (Job Vacancies)
• Application for Leave

9. City Mayor’s Office

• Releasing of outgoing communication
• medical/ financial/ burial / educational assistance

10. City Planning and Development Office

• Certifications/Certified Copies
• Simple Subdivision Approval (SSA) PreliminaryApproval for Locational Clearance (PALC) Development Permit (DP), Alteration ofPlan (AP) Application

11. City Treasurer’s Office

• Issuance of Certificate of Business Tax Retirement
• Payment of Business Tax / Real Property Tax and Fees/ Charges/Clearances
• Issuance of Community Tax Certificate or CTC
• Transfer tax assessment

12. General Services Office

• Procurement Services

13. Office of the City Civil Registrar

• Issuance of Certified True/Xerox copies of Civil Registry Records
• Registration of Birth
• Processing of Petitions for R.A 9048 and R.A. 10172(Change of First Name and Correction of Clerical Error in Civil RegistryRecords)

14. Sangguniang Panlungsod

• Application/Renewal of Mayors Permit
• Application for change of Certificate of Franchiseand order of Dropping and substitution of motor unit.

2. MESSAGES were classified as (a) LGU, (b) Public, and (c) Outside

Raw Data = 156

After Cleaning = 117

• Cleaning includes removing redundancy, URL/Links, mentions (@user1), and hashtags (#) from data

No. of datasets per Classification:

                            LGU = 74

                            Outside = 6

                            Public = 37

2.a MESSAGES FOR THE “LGU OF LEGAZPI”

Test Run was set to:

Topics = 5

No. of Iterations = 5

No. of Words = 5

Optimization Interval = 5

Topic Labels	Topic Models	Topic Distribution
0 Internet Connectivity	internet, facility, easy, transaction, tulong	20%
1 Power Interruption	work, transaction, kuneksyon, brownout, trabaho	20%
2 Unrecognized	good, serbisyo, parking, leader, nangangailangan	20%
3 Online Services/ Paperless Transactions	service, community, paperless, murusdut, access	20%
4 Service Delivery	serbisyo, inaasikaso, paraan, nanghihingi, madaling	20%

Details:

Topics = 5

Topic Bits = 3

Topic Mask = 111

Maximum Tokens = 7

Total Tokens = 83

Time: 30 seconds each topic model

Sample Messages Sent for each Topic Label:

Topic 0:  Internet Connectivity

• “bilis at serbisyo kaya lang mabagal ang internet”

Topic 1: Power Interruption

• “APEC lagi brownout”

Topic 2: Unrecognized

• “Paki gawan po ng paraan ung mga illegal parking along rizal st.”

Topic 3: Online Services/Paperless Transactions

• “improve internet facility for paper less transaction”

Topic 4: Service Delivery

• “wala chair sa may public assistant. sana mag provide sila ng chair. mejo magulo at crowded rin sa entrance”
• “gandaang serbisyo”
• “Go go go LegAZPI”
• “more power pagpatuloy ang ganda serbisyo public”
• “dapat pantay pantay ang pakikitungo sa mga citizen. may mga pinipili munangasikasuhin”
• “maging masigasig sa pag tanggap sa mga kliyente. kung pwede gawin na mamayaman muna”

Visualization of the “Message – LGU” Topic Models:

2.b MESSAGES SENT FOR THE “PUBLIC”

Test Run was set to:

Topics = 3

No. of Iterations = 5

No. of Words = 5

Optimization Interval = 5

Topic Labels	Topic Models	Topic Distribution
0 Public Utility Vehicles Route	driver, traffic, city, ayos, route	33.33%
1 Traffic concerns	traffic, apec, flow, counter, working	33.33%
2 Road Conditions	gamit, accident, ayos, prone, daraga	33.33%

Details:

Topics = 3

Topic Bits = 2

Topic Mask = 11

Maximum Tokens = 40

Total Tokens = 104

Time: 36.64 seconds each topic model

Sample Messages Sent for each Topic Label:

Topic 0: Public Utility Vehicles Route

• “Nakakahilong route ng jeep”
• “masasanay din sa bagong route”

Topic 1: Traffic Concerns

• “Lumaki yung traffic sa Washington”
• “sobra traffic!”
• “Grabe ang traffic sa SM”
• “so many unofficially designate parada of tricycle and colurum in addition sobra huminging pamasahe ang iba driver for jeep, cutting trip specialy Tahao Road route”

Topic 2: Road Conditions

• “Road Sign: one way or this way only to Lapu-Lapu street. Driver not from Legazpi city would counter flow traffic from Ayala mall going to Daraga. same with Peñaranda St., Driver from Bonot, Gogon would counter flow traffic from  St Raphael Church Asphalt road (overlapping)from Gaisano to UST/ Aquinas hospital is accident prone. personally witnessed a lot of near miss and actual accident Newly installed traffic lights are not working in some areas of Legazpi city. Thank you”

Visualization of the “Message – Public” Topic Models:

2.b MESSAGES SENT FOR “OUTSIDE”

Test Run was set to:

Topics = 3

No. of Iterations = 5

No. of Words = 4

Optimization Interval = 5

Topic Labels	Topic Models	Topic Distribution
0 Traffic Concerns	sasakyan, araw, kalsada, council	33.33%
1 Road/Street Obstruction	barangay, traffic, nakaparada, daraanan	33.33%
2 Barangay Document Issuances	hall, problema, lagda, dokyumento	33.33%

Details:

Topics = 3

Topic Bits = 2

Topic Mask = 11

Maximum Tokens = 31

Total Tokens = 41

Time: 36.64 seconds each topic model

Sample Messages Sent for each Topic Label:

Topic 0: Traffic Concerns

• “paki ayos ng traffic”

Topic 1: Road/Street Obstruction

• “Ang mga sasakyan sa tabi ng kalsada na nagiging problema para maging masikip angdaraanan ng ibang mga sasakyan. Ang daming nakaparada na sasakyan sa tabi ngkalsada.”

Topic 2: Barangay Document Issuances

• “Bilang isa miyembro ng syudad na to, inaasahan naming mga individual na nakatira sa barangay, na sana pagdating sa mga barangay council, ay andun sila sa Barangay Hall araw araw (8-5) o kahit isa barangay council sa bawat araw, kasi may mgatao na nangangailangan ng tulong nila. Isa pa, may mga pagkakataon naman na maymga kumukuha ng Barangay Clearance o mga dokyumento na kailangan ng lagda nilangunit wala sila sa Barangay Hall at tanging ang Barangay Secretary lang ang laging nandoon.”

Visualization of the “Message – Outside”  Topic Models:

Ms. Lany Maceda of Bicol University College of Science is currently finishing her Doctorate Degree in Information Technology at the University of the Cordilleras Baguio City.