At Bricolage we use cutting-edge text mining and data analytics technologies and research techniques, which allow us to combine different types and sources of data, then examine and analyse them, to discover useful knowledge and gain new insights.
We specialise in taking both structured and unstructured data and bringing them together to spot and extract patterns, correlations, anomalies, entities, topics, themes and facts that would not have been apparent had just one type of data been used.
By combining data in this way, it becomes possible not just to measure phenomena – to say what is happening and when – but also to look at their possible causes – to understand the ‘why’ behind the ‘what’ – and develop predictive models – to forecast future performance and outcomes.
Preparation of all sources of textual and numeric data to make them available for analysis.
Combining data of different types and from different sources to gain additional insights into operations and behaviours. For example:
Mixed method analysis of quantitative and qualitative data providing added insight, depth and causal understanding.
Analysis of operational/administrative data (such as CRM database records, spreadsheets, web feedback forms, questionnaires, surveys, reports, documents, financial records, and emails) for:
Extraction of patterns, correlations, trends and anomalies from any available data source.
Using visualisation tools to both explore and analyse data, and to display the results to create clarity and enhance understanding.
Rapid analysis, categorisation and annotation of large collections of all types of electronic text (including social media) to discover themes, topics, and sentiments.
Mining text data for information and facts contained within, including entities such as people, companies, products, places, events, and dates, and the relationships between them.
Enriching text data through the addition of related metadata.
Creating bespoke, industry specific taxonomies and dictionaries for use in Natural Language Processing (NLP), Machine Learning and automated analysis.
Developing statistical models to predict outcomes based on current and historical data in order to help plan and target future resources more effectively.
Carrying out detailed evaluation of the accuracy and consistency of data collection structures and CRM systems.
Developing and refining data collection, processing and delivery services, and user experience.
Evaluating your organisation’s use of data and the understanding of its potential uses.
Planning a data collection, user education and usage strategy to evolve data usage within the organisation.
Quantitative, Qualitative and Mixed Methods research design.
Designing data collection tools:
Project Management, facilitating workshops and delivering training.
Accessing additional specialists and domain experts:
A presentation that outlines the approaches taken to capturing and analysing logistics lessons from a corpus of 4 million words created from multiple natural language data sources for the HERRICK (Afghanistan) operations. A total of 767 lessons were extracted, covering 8 major thematic areas, which allowed for better forward planning of future operations.
Please click here to download this resource.
A downloadable document detailing services offered by Bricolage
Please click here to download this resource.
A report detailing how, during 2009-2010, some vulnerable and homeless clients of Wiltshire Council’s Housing Options Service were denied the statutory duties owed to them or were not offered good quality and timely housing advice. The report is based on case studies and a line by line text analysis of anonymised CRM database records. This mixed method approach forms a detailed picture of the treatment of individuals and groups presenting as homeless and points to weaknesses in Wiltshire Council’s procedures and practices for processing and supporting them.
Please click here to download this resource.
Ready to unlock the full potential of the data you hold?
