Content
What data do you provide?
A complete list of our data coverage by country, year, and kind can be found in our Data Coverage table. You can also use this table to see which countries have full-text data and images, as well as what languages are available for each country. The extent of coverage differs based on your subscription level.
Do you provide patent images?
Yes. We provide full-page PDFs, clipped images, drawings, chemistry files, and other non-textual files delivered by patent offices. Access to the images differs depending on your subscription level. We also offer a Bulk Attachment service which provides access to complete copies of recently published attachments packaged in a convenient archive. Consult the Data Coverage table to see what types of images are available for specific countries. For more information about the types of images available and how to access them, see Attachment Server Description and Coverage.
Which countries have IFI Snapshots including patent status, expiry dates, standardized names, and claims summaries?
IFI Snapshots are available for the US, EPO, Australia, Canada, Germany, Spain (without claims summaries), France, Great Britain, and Japan. Chinese Snapshots are available for Premium Plus subscribers only. Availability by country can be viewed in table form on the ifi-integrated-content page.
Do you have PAIR/PEDS data?
PAIR/PEDS data is available through the Legal Status service of our API. This service provides transaction histories from PEDS which, according to the USPTO, may lag behind Public PAIR by a few days. Since this service is currently BETA, request parameters and response formats may change.
Do you have design patents?
We receive some design patents from DOCDB so there are some records with bibliographic information in the database. However, in general, we do not cover design patents, and full text is not available for these records. A quick way to find which designs are available in CLAIMS Direct is to query by kind code since all of them share kind code "S". "E" is also used for design reissues. Therefore, you would search on pnkind:(S* OR E*). You can then see the country coverage for designs by faceting your query by patent country (pnctry). For more information about how to use faceting in CDWI, see the Facet page.
Do you have provisional patents?
US provisional patents are only available in CLAIMS Direct if they have been published by the USPTO. Provisional patents are generally not published, so they are rarely available in CLAIMS Direct. When a US provisional patent is transformed to a "normal" patent, then it is published. The new patent will have different publication and application numbers, but the application number of the provisional patent is always included in the priority data of any member of the family. This application number is the one with a kind code "P". You can find them by querying ankind:P AND anctry:US. Other countries may work differently. Some of them publish provisional patents, such as Australia. Also, the effect of a provisional patent may be different in different countries.
How long does it take for translations to become available?
Our typical delivery schedule, including the timing of translations, may be found on the Content Updates page. We are now providing English translations from Google Translate with the exception of China, which is receiving improved English translations from a separate source. This change has increased the speed with which translations are available. For all countries except China, translations are now provided one day after the availability of the original text. For Chinese records, English translations are available one day after the publication date.
Do you provide citations?
Yes. CLAIMS Direct provides forward and backward citations, including patent and non-patent literature citations. The patent citations may be accessed using our Citation Service. To access this service through an on-site instance, see Leveraging On-Site Citation and Family Functionality. Citations are also searchable in CDWI. Visit the Citations page to see which search fields are available. For more information about the citation data, see technical-data citations.
Do you provide patent family information?
Yes. Simple family IDs, as assigned by DOCDB, are available through both CDWI and CDWS using the search field fam. IFI CLAIMS assigns a default family ID of -1 if we have not received a family ID from DOCDB. If you see this value, it means either that the record is not available in DOCDB or we have not yet received the DOCDB record. In the latter case, the family ID will be populated once we receive the DOCDB record.
Extended families are only available through CDWS. This is because the extended family can expand literally daily so it must be generated as needed. Instructions on how to access family information through the API can be found on the Family page. To access family data through an on-site instance, see Leveraging On-Site Citation and Family Functionality.
How can I tell if a patent is still in force or what its current status is?
We get legal status data for EP patents from the EPO's subscription product, INPADOC. For China, Premium Plus subscribers also receive legal status data from our Chinese data source. This information is available in the legal-status container of the XML files.
We also provide USPTO PAIR data through the API's Legal Status service. This service is currently BETA so request parameters and response formats may change.
For countries which contain IFI Snapshots, IFI calculates current legal status. This includes anticipated expiration dates, information about regulatory extensions, terminal disclaimers, and more. All of this information can be found in the ifi-patent-status container in the XML files, which can be retrieved using the Text service. For more information, see our documentation on the ifi-integrated-content.
How do you calculate the IFI standardized form of the current assignee?
IFI curates the current assignee only for those countries for which Snapshots are available. To calculate current assignee, we use legal event data along with other sources of information. For the US and AU, we receive additional data from the patent offices, which can be found in the assignment history section of the XML. We also do some of our own rules-based work: to identify "individual" (e.g., inventor matches assignee), to standardize the names (e.g., “co” vs “company”), and to create the probable assignee on an application where one isn't present. Our editorial team also creates relationships between names, which are then used in the calculation of the ifi-name-current. For example, when Facebook became Meta Platforms, the change was quickly identified by our editors and the ifi-name-current field was updated to reflect the new name.
Do you make any changes to the data provided in the documents?
No, we do not correct or alter the data we receive from our providers except to normalize data from multiple sources into a consolidated and coherent output format and XML structure. We do report known errors to patent authorities but we do not make changes to the data unless we receive corrected records from our providers. The one exception is the IFI Snapshot data, which we calculate. IFI’s patent status and expiration indicators should not be considered authoritative. An attorney or the relevant patent office should be consulted for an authoritative expiration and status determination.
What does a publication or application date of 10101 or 19000101 mean?
If no date is present in the data we receive from our providers and a document-id element exists, a default value of 00010101 (which may appear as 10101) or 19000101 is inserted. This is done because a date value is required for our processing and to facilitate Solr searching. The 00010101 date is available in Solr responses only and is not present in the XML.
Installation
What options are available for installing CLAIMS Direct on-site?
If you want to install CLAIMS Direct on-site, we recommend three possible options: stand-alone monolithic, basic distributed, and advanced distributed installation. These are discussed on the Installation Types page.
Do you provide instructions for on-site installation?
Yes. Please see our Step-by-Step Installation Guide. If you are interested in our assistance with indexing the data warehouse, please contact your sales representative. For additional help, please contact us at support@ificlaims.com.
How can I install CLAIMS Direct using AWS (Amazon Web Services)?
Instructions for installing CLAIMS Direct using AWS can be found on the AWS Installation page.
Updates
How often is the data updated?
The CLAIMS Direct update schedule can be seen on the Content Updates page.
How are updates delivered?
With hosted services, you don’t need to worry about updates. It is easy to identify newly updated records via the API. For an on-site server, there is an update process that uses a web service to retrieve data updates that are published sequentially. The server downloads the new data, updates the XML repository, and triggers a Solr indexing process if you are running that server. There are a variety of load types, but all of this and more is explained in more detail on the Content Update Architecture page.
What tools are involved in the updates?
CLAIMS Direct uses the utilities apgupd and aidxd. See Update Utilities for more information.
For information about configuring a server to process updates, see the Processing Server page.
How can I tell which records have been updated?
For on-site instances, the modified_load_id indicates the last time a record was modified and therefore can be used to track updates. This topic is discussed in more detail in the blog post Sorting Through Data Warehouse Updates.
In the XML, there are two ways to tell if a document has been updated:
@statusattribute on thepatent-documentelement@file-reference-idattribute on thepatent-documentelement
If you want to see at a glance whether a document has been updated, use the @status attribute. If you are persisting the data, you should be storing the file-reference-id which can then be used to flag or recognize updates.
For example, this sample record has been updated a few times since publication:
| file-reference-id | load-source | Description |
|---|---|---|
| 303392 | IFI-Cont | ifi-integrated-content update |
| 303372 | DOCDB | DOCDB update |
| 302363 | USREAS | Reassignment update |
| 302181 | US | Initial load |
You can see that the file-reference-id of an update will have a larger integer than the original delivered value.
How can I tell if my on-site instance is in sync with CLAIMS Direct?
Run the following query on your on-site instance:
SELECT max(modified_load_id) FROM xml.t_patent_document_values;
Compare the results with the Last loadid shown on the cdwi.ificlaims.com home page.
The results do not need to be an exact match. Anything within approximately 100 load-ids is considered to be in-sync.
Another method is to query the System API and compare the completed_stamp member of the response.
Example:
wget -O- \ --header 'x-user:x' \ --header 'x-password:y' \ 'https://cdws21.ificlaims.com/system/dwh/loading/info?load_id=618115'
Response JSON:
{
"time" : "0.108006",
"status" : "success",
"load" : {
"load_source" : "IFI-Cont",
"completed_stamp" : "2022-04-13 04:33:21.378294",
"load_id" : 618115,
"running_status" : "complete",
"completed_status" : "success",
"entered_stamp" : "2022-04-13 04:33:04.767529",
"client_load_process_id" : 232605,
"url" : "https://alexandria.fairviewresearch.com/dwhpp/20220413/000000618115-IFI-Cont.tar.gz",
"ndocs" : 1097
}
}
Are updates received only for active patents or is it possible to receive an update for any patent?
Most updates affect patents that haven't expired yet or have expired recently, but there are some exceptions. For example, if there is a change in a CPC class that affects records related to a specific class code, then all records with that class code will be updated, whether they are expired or not.
Using CLAIMS Direct
What web services are available from CLAIMS Direct?
The CLAIMS Direct Web Services provide a comprehensive API that allows for search, XML data retrieval, reporting, specialized services for family and citation lookups, and more. For a full listing and description of the services available, see the Developer's Guide.
How can I access the full text?
The Text service is used to retrieve XML data from the CLAIMS Direct Data Warehouse.
To access the full text through SQL, use the following command:
SELECT xml.f_patent_document( xml.f_ucid2id(‘US-5551212-A’), null );
The full text can also be viewed using CDWI (see Viewing the XML Record).
How can I access patent images?
Attachments are housed on a specific attachment server. You may retrieve them through the Attachments service using our API.
To get PDFs, you can make this call with just the ucid:
wget -O- \ |
To get the first drawing (the image that normally appears with the abstract), make this abimage call:
|
Note that not all documents have PDFs or abstract images, so your application should check for errors.
To get all of the attachments, you can request the full list of available attachments:
|
This will return a list (in JSON) of the available attachments, with links to fetch them. The file names should correspond with the figure references in the document.
You can also see PDFs and attachments in CDWI. Do a search and produce a list of documents. Right click on a document and render "PDF" or “Attachments”. For attachments, you will see a list of available attachments which you can click to open. For biological and chemical patents, there are sometimes .MOL (molecule) and .SEQ (genetic sequence) files.
We also offer a Bulk Attachment service which provides access to complete copies of recently published attachments packaged in a convenient archive.
Some images appear to be missing from CLAIMS Direct. Is there a way to access them?
Japanese records frequently contain multi-page TIFFs. Multi-page TIFFs are indicated when multiple id attributes point to the same file. Many TIFF readers do not read any image data beyond the first IFD (image file directory). Therefore, you may see only the first page.
For example, 2007280735.tif is a multi-page TIFF containing the front page (000001), the abstract drawing (000002), and the drawing sheets (000003-000012) referenced in the drawings section of the XML:
$ tiffinfo 2007280735.tif |grep ImageDescription ImageDescription: 000001 # front page ImageDescription: 000002 # abstract drawing ImageDescription: 000003 # drawing sheets ImageDescription: 000004 ImageDescription: 000005 ImageDescription: 000006 ImageDescription: 000007 ImageDescription: 000008 ImageDescription: 000009 ImageDescription: 000010 ImageDescription: 000011 ImageDescription: 000012
The XML for this image shows that the file attributes point to the same physical file (2007280735.tif):
<img id="000003" he="101" wi="75" file="2007280735.tif" img-format="tif" img-content="drawing"/> <img id="000004" he="65" wi="70" file="2007280735.tif" img-format="tif" img-content="drawing"/> <!-- etc … -> <img id="000012" he="33" wi="53" file="2007280735.tif" img-format="tif" img-content="drawing"/>
A list of viewers providing proper handling of multi-page TIFFs can be found here.
Can I see the XML DTD and schema?
Yes. The DTD and schema are available here.
How do I create a report?
Our Reporting service allows you to create, process, and retrieve either CSV extracts or XML extracts from CLAIMS Direct.
Reports can also be created through CDWI. See Report Options for instructions.
Are there limits on the API services?
IFI CLAIMS introduced shared-service quotas beginning in the third quarter of 2019. Rate limits for shared services vary based on your subscription and service levels although the standard rate limit for most shared services ranges from 60-240 request units per window per client login, where window can range from 1 minute to 1 day. You can read more about these quotas here. In addition, there are some limits applied to the individual services. All of the API services which take multiple ucids as parameters have a limit of 100 documents per batch. That number provides a good balance between time to prepare the extract and time to push it across the network. The Search and Reporting services are limited to returning 30K documents per batch because deep paging can place undue stress on the system.
Is there any time when CLAIMS Direct's services are not available?
We have a regular maintenance window on Saturdays from 12:00-4:00pm Eastern Standard Time during which there may be brief service interruptions. If we have extended outages, a notification will appear on the Service Notifications page.
Searching
What search rules and syntax are used in CLAIMS Direct?
The CLAIMS Direct index is created and maintained using Apache Solr (http://lucene.apache.org/solr/) and hence, based on a schema definition, all of the functionality provided by Solr is available through CLAIMS Direct, including the entire Apache Lucene query language, result set sorting, facet searching, grouping etc. For more information and examples, see the Solr Searching Guide.
The basic search syntax used in CLAIMS Direct is Solr field name:value. For example:
cpc:a61k |
You can find more information about how to search using the API in the Developer's Guide.
For instructions on how to search using the CLAIMS Direct Web Interface (CDWI), see the CDWI Guide.
What fields are available for searching?
Fields available for searching in CLAIMS Direct, along with tips and examples, can be seen in the Solr Search Fields section. The CLAIMS Direct Solr index does not store all text. This means that although the entire index is searchable, only certain fields are available for inclusion in a result set. Retrievable fields are marked as "stored" in the Solr Search Fields tables and a quick reference list is available in the Quick Guide to Stored Solr Fields. You can also find them in the schema where @stored="true", or use the schema method of the Search service to inspect field attributes.
How do I search for a particular patent number?
There are several ways to search for a particular patent number in CLAIMS Direct. Use the an field to search for an application number, and pn to search for a patent number. You may also search for a ucid, which is a unique document identifier comprised of the country, document number, and kind code, e.g., US-96142365-A.
Number formats in CLAIMS Direct are generally provided in DOCDB-normalized format with a few exceptions, which are:
WO: WIPO's format has changed over the years and you can find 3 formats in DOCDB. We have standardized WO publication numbers to WO-YYYYNNNNNN. For example, the DOCDB number WO8002869 becomes WO-1980002869-A1 in CLAIMS Direct.
US applications: The DOCDB format is YYYYNNNNNN, while the CLAIMS Direct format is US YYYYNNNNNNN.
Patent document numbering is a complex subject, however, so you should review the following pages for more detailed information: application-reference and publication-reference. We also recommend checking the Release Notes for patent number or kind code changes.
If you have a list of numbers to search, you may use the List search option in CDWI, which you can read about here. This option includes a validation feature which converts some patent publication numbers into CLAIMS Direct's standard ucid format.
Before you conduct any searches, we recommend that you familiarize yourself with the searching options available in CLAIMS Direct using the Solr Searching Guide.
How do I search for a particular class or subclass?
CLAIMS Direct allows searching for Collective Patent Classification (CPC), International Patent Classification (IPC and IPCR), FI classification/F-Terms, US Classification, and European Classification (ECLA).
Our system allows searching by level. For IC, CPC, FI and ECLA, the different levels of classifications are indexed as follows:
l1=section, e.g., A
l2=section+class, e.g., A61
l3=section+class+subclass, e.g., A61K
l4=section+class+subclass+group, e.g., A61K0001
l5=section+class+subclass+group+subgroup, e.g., A61K0001001
For US classification, the different levels are indexed as follows:
l1=class, e.g., 161
l2=class+subclass (as integral number), e.g., 161234
l3=class+subclass with trailing decimal if exists, e.g., 161234 [.xxx] -->
For example, if you wanted to search for everything in CPC class A61K 9/2893 , the most efficient method would be to search by CPC level 5: cpcl5:A61K00092893
The search fields are listed, along with examples, on the Classifications page. More information about the classification data can be found here.
Before you conduct any searches, we recommend that you familiarize yourself with the searching options available in CLAIMS Direct using the Solr Searching Guide.
How do I search for a particular applicant or assignee?
There are several fields you can use to search for an applicant or assignee. See Associated Parties for a full list of searchable parties. Many of these fields are available in various formats, including the original format as filed by the applicant and the EPO’s standardized format. More information about formats can be found on the Common XML Attributes page.
A particularly helpful field to be aware of is the pa field, which provides a comprehensive search of the applicant, assignee, and reassignee fields. We also offer search fields like asgfull which allow you to search the name and address fields together, which may help to distinguish entities with the same name in different locations.
For countries which have IFI Snapshots, we offer standardized names. As you can see on the IFI Snapshots page, we provide standardized names for the original and current assignees. Similar to the pa field, there is an ifi_pa field which provides a comprehensive search of the original and current standardized assignees. Additional information about our standardized names can be found on the ifi-integrated-content page.
Before you conduct any searches, we recommend that you familiarize yourself with the searching options available in CLAIMS Direct using the Solr Searching Guide.
Contact Us
Still need help? Please send your questions, comments, and suggestions to support@ificlaims.com.