Finished projects
Back

Tracking Barcelona Brand in digital media.

Project's description and mission:

In relation to the Barcelona Brand project driven by Barcelona Global, Accenture´s proposal would allow us to create a tool to track the Barcelona Brand, which in turn would be useful for:

  • The creation of a platform that permits us to monitor the reputation of Barcelona as a brand in the online world (Twitter, Forums, Online Press, Corporate Communication).
  • With the information collected, we will be able to pinpoint and work specifically on priority areas to proactively manage the city’s image.

The model has scalable parameters, information sources, and comparability with other cities of reference. 

Specific objectives of the project:

  • Monitoring the Barcelona brand in digital media is necessary for continuously measuring Barcelona Global’s achievement towards the goals of Barcelona Global 
  • Analysing attitudes and feelings, by means of a combination of Big Data and Analytics, could improve decision-making, allowing Barcelona Global to detect the opinion reflected in digital media in the sectors of activity of the association’s interest
  • This monitoring could be used for later data comparison (public or generated) from other cities; tracking and comparison of brand Barcelona’s development over time and analysis of the effect of different initiatives carried out by the platform or the city (media campaigns, relevant events, etc.)

Proponents Partners or partner and team driver:

Raimon Miret (raimon.miret@accenture.com) - Products Managing Director

Belén Villasante Corredoira (belen.villasante@accenture.com) - Managing Director

Project description:

APPROACH

To achieve these objectives, Accenture proposes an implementation project whose outcome will be an ongoing service operated from Accenture’s facilities. The project will contain three stages:

1. Filter, cleanse, and prepare the input data with news and messages from digital media to be ready for use in analytic models. The infrastructure must be prepared to extract large volumes (Big Data).

2. Use of supervised models for text mining to automatically classify documents by topic and attitude (positive, negative, neutral). The topics to monitor will be the three agreed on in the recent meetings as well as the selected hot topics. These will be obtained in the pilot phase.

3. Construction of significant metrics (scope and degree of agreement) to track the impact of the initiatives carried out.

FINAL SOLUTION

Once the two phases of the project are implemented (pilot and setup), the workflow of the service will be the following:

  1. Identify filters to apply
  2. Extract, process and validate data
  3. Obtain Hot Topics in the pilot phase and select final topics to monitor
  4. Design, train, and execute ongoing models to classify by topic and analyse by attitude using Aqua Text Mining
  5. Enrich database with classification of documents
  6. Track and write reports adding information to document level in relevant metrics.

PILOT PHASE

Test concept in reduced time where, manually using just one data source (Twitter), the viability of monitoring topics that interest Barcelona Global can be checked (up to 10 topics).

  • Technology: Laboratory
  • Sources: Twitter
  • Text categorisation: Manual, for just one city
  • Results:
  • Manual categorisation of Twitter messages
  • 1 results report (PPT): Hot Topics, attitude, development over time and metrics

METRICS

  • Conclusions of viability and scope of ongoing service implementation

ONGOING SERVICE

Complete implementation project of a service (aaS) based at Accenture, where continuously and automatically the topics determined in the pilot from sources, provided by an external company, are categorised and valued as positive, negative, or neutral.

  • Technology: Big Data at AIO installations (Accenture Infrastructure Outsourcing)
  • Sources: Twitter + Moveover (Press and Forums, Facebook initially excluded)
  • Text categorisation: Automatic via Aqua Text Mining (Accenture’s asset) for one city (*)
  • Results:
  • Automatic real-time categorisation of documents
  • Reports: trending topics, attitude, development over time and metrics, with monthly refresh

Model performance evaluation. Frequency to be confirmed.

Expected results:

PILOT PHASE

  • Manual categorisation of Twitter messages
  • 1 results report (PPT): Hot Topics, attitude, development over time, and metrics
  • Conclusions of viability and scope of on-going service implementation

ONGOING SERVICE

  • Automatic real-time categorisation of documents
  • Reports: trending topics, attitude, development over time and metrics, with monthly refresh

Model performance evaluation. Frequency to be confirmed

Back

This site uses own and third-party cookies to provide a better experience and service. When browsing our website you agree to our use of cookies.

Accept