Leximancer is an automated text-mining tool produced by an Australian company of the same name, and I recently had the opportunity to test an online version of the tool called LexiPortal. Leximancer creates detailed lists of concepts and tags prevalent in a set of documents or data, and the user can then customize the concept lists and create a “concept map” or generate a report through the Insight Dashboard. Although I experienced some problems with uploading documents in the trial version of this system, the tool still may be very useful to patent analysts who want to identify common themes within a set of patent documents.
After the jump, learn about the different versions of Leximancer, how to access the tool, and read about my experience testing the trial version of LexiPortal!
According to the Leximancer website, versions of the Leximancer software include:
- Leximancer Enterprise – A multi-user server configuration that runs within a desktop browser. Capabilities of this version include:
- Use of Leximancer’s web-crawling and search features to extract content directly from the internet.
- Publish “view-only” maps to share around the organization for evaluation and analysis.
- Leximancer Portal (LexiPortal) – A service available via any supported web browser in a hosted “on-demand” platform.
- Leximancer Desktop – Service on stand-alone personal computer or laptop.
For the LexiPortal service, users can subscribe in a “pay-as-you-go” environment, where they “do not have to subscribe every month, but are charged only for your actual months of usage – on either a time or service basis” (according to the website). Separate monthly rates are charged for the following services listed on the Leximancer website:
- Leximancer “internal” data analytics – Upload internal data files (including but not limited to doc, rtf, csv, txt, pdf and html files).
- Web-crawling “external” data acquisition – Fetch web data from specific sources via a URL (with optional keyword content filtering or user chosen web search terms).
- Leximancer Insight Dashboard service – Users can view Dashboard reports generated by a Leximancer consultant.
Users can also request also request a free trial for LexiPortal through the Leximancer website.
After signing in to LexiPortal, the user can select to create a new project or work on a current project through the side menu “Manage Projects”, which contains a hierarchical list of folders and projects. By right clicking on a folder, the user can select to create a new folder (within the current folder) or create a new project.
After naming the new project and entering a description, a new “Project Control” menu will appear in the main window. Multiple projects can be open at one time, each in a separate tab within the main window.
Within the project control menu, the user must complete each step listed on the menu, until all steps are listed with a “Ready” status:
- Load Data – Select data currently uploaded to the system (add to “selected data” section), or upload new files to add to the “selected data” section. I was unable to upload a collection of patent PDF documents to the system through my trial account, so I had to run the concept analysis on data sets already uploaded to the system. It is unclear whether this inability to upload a set of PDF documents was due to the trial status of my account on LexiPortal or a temporary glitch in the system.
- Generate Concept Seeds – Before selecting to run this section, edit the following settings (display options through “Show Settings” button):
- Text Processing Settings – Define general (sentences per block, prose test threshold, etc.), tagging (apply folder tags, apply file tags, etc.), and performance (file preparation and processed file storage format) settings.
- Concept Seeds Settings – Select general concept seeds identification options (automatically identify concepts, total number of concepts, etc).
- Generate Thesaurus – Before selecting to run this section, edit the following settings (display options through “Show Settings” button):
- Concept Seeds – View lists of auto concepts, auto tags, user defined concepts, and user defined tags. Select to add, edit, remove, download, upload, and merge/unmerge concepts. User defined concepts and tags have the additional option of “sentiment lens” (add sentiment concept seeds and compound concepts).
- Thesaurus Settings – Define general (learn concept thesaurus using source document, concept generality, learn from tags, etc.) and concept profiling (number to discover, themed discovery, etc.) settings.
- Run Project – Before selecting to run this final section, edit the following settings (display options through “Show Settings” button):
- Compound Concepts – View lists of name-like concepts and word-like concepts, and merge selected concepts using Boolean operators (AND/OR/NOT) to create compound concepts.
- Concept Coding Settings – Select from lists of general concept types, available names, and available concepts to define mapping concepts, kill concepts, and required concepts. Define additional options (word-like concept classification threshold, name-like concept classification threshold, etc.).
- Project Output Settings – Define options of Concept Map (map type, size), and Insight Dashboard (general settings, quadrant report settings, and lists of categories and attributes).
After uploading the data and running the “Generate Concept Seeds” and “Generate Thesaurus” steps, the user can select to “run project” to complete the analysis process. The user can always upload additional data, edit settings, and re-run the project to update the analysis.
After all statuses for the sections of the project control menu are listed as “Ready”, the user can select from the three options at the bottom of the menu:
- Concept Map – View a concept map in one panel and a tabbed selection of lists and navigation tools in a side menu: themes, concepts, thesaurus, pathway, and query (search option). The user can customize the appearance of the map through a variety of options in a horizontal menu above the map, and the map can be exported or saved.
- Insight Dashboard – A report available in PDF, HTML, or CSV formats and divided into five sections (section definitions from a generated Insight Dashboard report):
- Quadrant Overview (“a high-level, visual chart displayed in a ‘magic quadrant’ format”)
- Ranked Concepts for Categories Overview (“a more quantitative analysis in ranked barchart format of the most prominent Concepts within the particular Category”)
- Ranked Compound Concepts for Categories Overview (“similar to Section 2, providing a ranked list of the most prominent Concept pairs for the Category”)
- Supporting Text Overview (“This provides a good supporting text excerpt, and real evidence, for the top related Concepts – for each of the Concepts identified and ranked in Section 2.”)
- Ranked Concept Count (“This provides the actual ranked list of ALL Concepts and their associated reference count from the original base data.”)
- Data Exports – The pop-up window for data exports lists the following available exportable reports:
- Pairs of Concepts across the entire Data Corpus
- List or pairs of Concepts within each Text Excerpt
- List of Concepts within each Context Block, within each Text Excerpt
- Sentiment Lens Seed Set
- Pairs of Concepts across the entire data corpus (frequency count)
- Pairs of Concepts across the entire data corpus (inverted prominence)
The user can either mouse over various options to view further descriptions or explanations, or the user can select the option “Help” in the upper left hand corner, where they can select to view a PDF version of the Leximancer manual.
If a user is able to upload any set of documents or data to LexiPortal and create a customizable list of concepts and associated visualizations, then this tool may me useful for patent analysts looking for common themes in a set of patent documents or non-patent literature documents. It is unclear whether my own inability to upload a set of patent documents to the system was due to a mistake in my use of the system, a temporary system glitch, or the trial status of my account. I would recommend testing the capabilities of the tool yourself by requesting a free trial for LexiPortal on the Leximancer website, to see if the tool is useful for your particular patent or non-patent literature analysis needs.
Have you used Leximancer before? What did you think of the service? Let us know in the comments!
Landon IP provides professional patent analysis services, including patent landscape studies. The patent landscape study is a comprehensive analysis of patents to benefit product managers, scientists, and patent attorneys. It is a map and a discussion of patent activity in a specific technology area that significantly improves your ability to make major business decisions.
This post was contributed by Joelle Mornini. The Intellogist blog is provided for free by Intellogist’s parent company Landon IP, a major provider of patent searches, trademark searches, technical translations, and information retrieval services.