QWAM Text Analytics

QWAM Text Analytics is the outcome of many years of R&D in semantics and artificial intelligence by QWAM teams in partnership with several well-known research and academic institutions. The QWAM Text Analytics product aggregates all of our expertise in content analytics and metadata extraction from textual data. It is aimed at content managers, content providers and editors,marketing professionals, technical writers, web intelligence specialists, and IT professionals in charge of content management applications. Its main applications are:

  • Sales, Market and Business Intelligence
  • E-reputation
  • Human Resources content, analysis of employee’s surveys and feedbacks
  • Analysis of clients verbatims and feedbacks
  • Media and publishing (SEO, contextualisation, tags generation…)

Unstructured data and textual content from a business perspective

Large-scale use of the internet and digital transformation have led to a massive amount of textual data, most of which is unstructured content ; these are essentially found

  • on the web : web pages, press articles, blogs and comments, client’s verbatims, posts on social networks, etc.
  • within companies and organizations : reports, studies, regulatory documents, contracts, internal online surveys, human resources content, ...

QWAM has heavily invested in the development and improvement of QWAM Text Analytics, especially with regards to Artificial Intelligence R&D programs. Companies using new technologies to extract and analyze textual data face many business challenges. These are the same across all industries and services and are an important part of their igital transformation and performance. QWAM helps them understand these challenges and bring solutions to their business case.

QWAM Text Analytics – What does it do?

Today, communication is mainly done through digital media, be they online surveys, customer feedback on blogs and forums, or employee posts on social networks or intranets. QWAM Text Analytics can help automate processes and treatment of textual content for different business contexts. The sentiment analysis module precisely qualifies the tone and subject of analyzed data.

The digital transformation of organizations often starts with a more general use of digital content, which mostly consists of documents. Thanks to its extraction module, QWAM Text Analytics can efficiently define each document with calculated metadata. As each metadata is considered a facet, it is therefore easy to build a “faceted search engine.” Document searches are done progressively. Starting with a simple keyword search, the user can focus on a sub-group of documents that meet certain criterion (facets), and then refine his search. Metadata also enables the automated classification of documents in folders or topics trees. With QWAM Text Analytics, organizations can put in place such processes by personalizing automatically generated metadata that correspond to business requirements and user needs.

For large sets of documents, QWAM Text Analytics can group documents based on specific criteria (client needs) and extract key information and insights. It is then possible to build new data based on extractions made by QWAM Text Analytics, such as identifying the relations between extracted entities, which can show how a person is linked to several companies. Texts are then organized as coherent subsets as per business needs, and can be accessed via dashboards or search engine.

QWAM Text Analytics - How does it work?

Three main components run the necessary processes and algorithms in QWAM Text Analytics:

  • First, the extraction engine, which identifies specific rules and treatments for company names, persons, concepts, products, events, etc.
  • Secondly, the discovery engine organizes semantic enrichment metadata by typology (business concepts, finance, politics, health, etc.). It also detects the appropriateness of sentiments with their nature (joy, anger, etc.), force (like, love, adore) and polarity (positive, negative), and reveals hot topics covered on the web that relate to your areas of interest.
  • Thirdly, the screening engine enables organizations to monitor the results of the first two modules and to make improvements where necessary.

General functional architecture of QWAM Text Analytics


Eric Roustit

Responsable Informatique Editoriale - Le Figaro
When faced with an increasing number of articles that needed to be processed, it became clear that we needed to automate semantic content enrichment and tag generation. QWAM has supplied us an efficient tool to do just this. We can now link to every article a list of relevant terms and recommend other articles on related subjects.

Jacques Sallou

Capitaine de corvette - Marine Nationale
Le bien-être et les conditions de vie et de travail sont au cœur des préoccupations. Dans l’armée c’est encore plus vrai, car les militaires risquent leur vie pour notre pays etleur bon moral est garant de la réussite de leurs missions. Nous avons donc le devoir de leur offrir les meilleures conditions de travail possibles. Pour cela, les retours qu’ils nous font par le biais des différentes enquêtes (telles que l’i2M, l’indicateur de mesure du moral) et des sondages sont soumis à des études poussées. Notre rôle est de fournir à nos décideurs une cartographie précise des attentes des marins, en quantifiant la force de ces préoccupations. Ils peuvent ainsi prendre les mesures nécessaires à l’amélioration des conditions d’exercice de leur métier. […]

Discover QWAM Text Analytics in details