System for ensuring the internal consistency of a fact repository

Description

RELATED APPLICATIONS

This application is related to the following applications, each of which is hereby incorporated by reference:

U.S. patent application Ser. No. 11/097,688, “Corroborating Facts Extracted from Multiple Sources,” filed on Mar. 31, 2005;

U.S. patent application Ser. No. 11/097,690, “Selecting the Best Answer to a Fact Query from Among a Set of Potential Answers,” filed on Mar. 31, 2005;

U.S. patent application Ser. No. 11/097,689, “User Interface for Facts Query Engine with Snippets from Information Sources that Include Query Terms and Answer Terms,” filed on Mar. 31, 2005;

U.S. patent application Ser. No. 11/142,853, “Learning Facts from Semi-Structured Text,” filed on May 31, 2005;

U.S. patent application Ser. No. 11/142,740, “Merging Objects in a Facts Database,” filed on May 31, 2005; and

U.S. patent application Ser. No. 11/142,765, “Identifying the Unifying Subject of a Set of Facts,” filed on May 31, 2005.

TECHNICAL FIELD

The disclosed embodiments relate generally to fact databases. More particularly, the disclosed embodiments relate to methods and systems for maintaining the internal consistency of a fact database.

BACKGROUND

The World Wide Web (also known as the “Web”) and the web pages within the Web are a vast source of factual information. Users may look to web pages to get answers to factual questions, such as “what is the capital of Poland” or “what is the birth date of George Washington.” The factual information included in web pages may be extracted and stored in a fact database.

A fact database may, at times, become internally inconsistent. When a fact database is populated with data, there may be gaps in the data for which the database building module does not have the data to fill. When fact database maintenance operations are performed, data may be modified or removed, resulting in possible data inconsistencies. These internal inconsistencies may diminish the quality of the fact database.

SUMMARY

According to an aspect of the invention, a method of improving internal consistency of a database includes accessing a set of objects in the database (e.g., a fact repository), each object including one or more attribute-value pairs, wherein at least a subset of the values in the attribute-value pairs includes respective links to other objects; filtering the attribute-value pairs of the objects to remove attribute-value pairs meeting predefined criteria, wherein objects meeting a null information criterion after the filtering comprise null objects; identifying attribute-value pairs of the objects including links to null objects; and removing the links to null objects.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 illustrates a network, according to some embodiments of the invention.

FIG. 2 is a block diagram illustrating a data structure for an object within a fact repository, according to some embodiments of the invention.

FIG. 3 is a flow diagram illustrating a process for dereferencing attribute-value pairs, according to some embodiments of the invention.

FIG. 4 is a flow diagram illustrating a process for checking references in attribute-value pairs, according to some embodiments of the invention.

FIG. 5 illustrates a dereferencing and reference checking system, according to some embodiments of the invention.

Like reference numerals refer to corresponding parts throughout the drawings.

DESCRIPTION OF EMBODIMENTS

Within a fact repository organized based on objects (representing entities and concepts) and facts associated with objects, a fact may reference another object. In other words, facts may serve as connections between objects. For example, the fact that Tokyo is the capital of Japan connects objects representing Tokyo and Japan. The reference to the other object may include a link to the other object (such as an object identifier) and a name of the other object. However, the name of the other object may be missing or incorrect, even though the link may be correct. Furthermore, the other object may be “removed” from the fact database during fact database maintenance operations, resulting in a dangling link. Both missing or incorrect object names and dangling links represent internal inconsistencies in the fact database. The internal inconsistencies may be remedied by inserting the name of an object into facts that link to the object and removing dangling links.

FIG. 1 illustrates a network 100, according to some embodiments of the invention. Network 100 includes one or more document hosts 102 and a fact repository engine 106. The network 100 also includes one or more networks 104 that couple these components.

The document hosts 102 store documents and provide access to documents. A document may be any machine-readable data including any combination of text, graphics, multimedia content, etc. In some embodiments, a document may be a combination of text, graphics and possible other forms of information written in the Hypertext Markup Language (HTML), i.e., a web page. A document may include one or more hyperlinks to other documents. A document may include one or more facts within its contents. A document stored in a document host 102 may be located and/or identified by a Uniform Resource Locator (URL), or Web address, or any other appropriate form of identification and/or location.

The fact repository engine 106 includes an importer 108, a repository manager 110, a fact index 112, and a fact repository 114. The importer 108 extracts factual information from documents stored on document hosts 102. The importer 108 analyzes the contents of the documents stored in document host 102, determines if the contents include factual information and the subject or subjects with which the factual information are associated, and extracts any available factual information within the contents.

The repository manager 110 processes facts extracted by the importer 108. The repository manager 110 builds and manages the fact repository 114 and the fact index 112. The repository manager 110 receives facts extracted by the importer 108 and stores them in the fact repository 114. The repository manager 110 may also perform operations on facts in the fact repository 114 to “clean up” the data within the fact repository 114. For example, the repository manager 110 may look through the fact repository 114 to find duplicate facts (that is, facts that convey the exact same factual information) and merge them. The repository manager 110 may also normalize facts into standard formats. The repository manager 110 may also remove unwanted facts from the fact repository 114, such as facts related to pornographic content.

The fact repository 114 stores factual information extracted from a plurality of documents that are located on the document hosts 102. In other words, the fact repository 114 is a database of factual information. A document from which a particular fact may be extracted is a source document (or “source”) of that particular fact. In other words, a source of a fact includes that fact within its contents. Source documents may include, without limitation, Web pages. Within the fact repository 114, entities, concepts, and the like for which the fact repository 114 may have factual information stored are represented by objects. An object may have one or more facts associated with it. Within each object, each fact associated with the object is stored as an attribute-value pair. Each fact also includes a list of source documents that include the fact within its contents and from which the fact was extracted. Further details about objects and facts in the fact repository are described below, in relation to FIG. 2.

The fact index 112 provides an index to the fact repository 114 and facilitates efficient lookup of information in the fact repository 114. The fact index 112 may index the fact repository 114 based on one or more parameters. For example, the fact index 112 may have an index that maps unique terms (e.g., words, numbers and the like) to records or locations within the fact repository 114. More specifically, the fact index 112 may include entries mapping every term in every object name, fact attribute and fact value of the fact repository to records or locations within the fact repository.

It should be appreciated that each of the components of the fact repository engine 106 may be distributed over multiple computers. For example, the fact repository 114 may be deployed over N servers, with a mapping function such as the “modulo N” function being used to determine which facts are stored in each of the N servers. Similarly, the fact index 112 may be distributed over multiple servers, and the importer 108 and repository manager 110 may each be distributed over multiple computers. However, for convenience of explanation, we will discuss the components of the fact repository engine 106 as though they were implemented on a single computer.

FIG. 2 illustrates an exemplary data structure for an object within the fact repository 114, according to some embodiments of the invention. As described above, the fact repository 114 includes objects, each of which may include one or more facts. Each object 200 includes a unique identifier, such as the object ID 202. The object 200 includes one or more facts 204. Each fact 204 includes a unique identifier for that fact, such as a fact ID 210. Each fact 204 includes an attribute 212 and a value 214. For example, facts included in an object representing George Washington may include facts having attributes of “date of birth” and “date of death,” and the values of these facts would be the actual date of birth and date of death, respectively. A fact 204 may include a link 216 to another object, which is the object identifier, such as the object ID 202 of another object within the fact repository 114. The link 216 allows objects to have facts whose values are other objects. For example, for an object “United States,” there may be a fact with the attribute “president” whose value is “George W. Bush,”, with “George W. Bush” being another object in the fact repository 114. In some embodiments, the value field 214 stores the name of the linked object and the link 216 stores the object identifier of the linked object. In some other embodiments, facts 204 do not include a link field 216 because the value 214 of a fact 204 may store a link to another object.

Each fact 204 also may include one or more metrics 218. The metrics may provide indications of the quality of the fact. In some embodiments, the metrics include a confidence level and an importance level. The confidence level indicates the likelihood that the fact is correct. The importance level indicates the relevance of the fact to the object, compared to other facts for the same object. The importance level may optionally be viewed as a measure of how vital a fact is to an understanding of the entity or concept represented by the object.

Each fact 204 includes a list of sources 220 that include the fact and from which the fact was extracted. Each source may be identified by a Uniform Resource Locator (URL), or Web address, or any other appropriate form of identification and/or location, such as a unique document identifier.

In some embodiments, some facts may include an agent field 222 that identifies the module that extracted the fact. For example, the agent may be a specialized module that extracts facts from a specific source (e.g., the pages of a particular web site, or family of web sites) or type of source (e.g., web pages that present factual information in tabular form), or a module that extracts facts from free text in documents throughout the Web, and so forth.

In some embodiments, an object 200 may have one or more specialized facts, such as a name fact 206 and a property fact 208. A name fact 206 is a fact that conveys a name for the entity or concept represented by the object 200. For example, for an object representing the country Spain, there may be a fact conveying the name of the object as “Spain.” A name fact 206, being a special instance of a general fact 204, includes the same parameters as any other fact 204; it has an attribute, a value, a fact ID, metrics, sources, etc. The attribute 224 of a name fact 206 indicates that the fact is a name fact, and the value is the actual name. The name may be a string of characters. An object 200 may have one or more name facts, as many entities or concepts can have more than one name. For example, an object representing Spain may have name facts conveying the country's common name “Spain” and the official name “Kingdom of Spain.” As another example, an object representing the U.S. Patent and Trademark Office may have name facts conveying the agency's acronyms “PTO” and “USPTO” and the official name “United States Patent and Trademark Office.” If an object does have more than one name fact, one of the name facts may be designated as a primary name and other name facts may be designated as secondary names.

A property fact 208 is a fact that conveys a statement about the entity or concept represented by the object 200 that may be of interest. For example, for the object representing Spain, a property fact may convey that Spain is a country in Europe. A property fact 208, being a special instance of a general fact 204, also includes the same parameters (such as attribute, value, fact ID, etc.) as other facts 204. The attribute field 226 of a property fact 208 indicates that the fact is a property fact, and the value field is a string of text that conveys the statement of interest. For example, for the object representing Spain, the value of a property fact may be the text string “is a country in Europe.” Some objects 200 may have one or more property facts while other objects may have no property facts.

It should be appreciated that the data structure illustrated in FIG. 2 and described above is merely exemplary. The data structure of the fact repository 114 may take on other forms. Other fields may be included in facts and some of the fields described above may be omitted. Additionally, each object may have additional special facts aside from name facts and property facts, such as facts conveying a type or category (for example, person, place, movie, actor, organization, etc.) for categorizing the entity or concept represented by the object. In some embodiments, an object's name(s) and/or properties may be represented by special records that have a different format than the general facts records 204 associated with the attribute-value pairs of an object.

An object is a collection of facts. An object may become a null or empty object when facts are removed from the object. In some embodiments, a null object is an object that has had all of its facts (including name facts) removed, leaving the object with only its object ID. In some other embodiments, a null object is an object that has all of its facts other than name facts removed, leaving the object with its object ID and name facts. In further other embodiments, where an object has names in special records that have a different format from general facts, the object is a null object only if all of its associated facts, not including the special records for its names, are removed. Alternatively, the object may be a null object only if all of its facts and the special records for its names are removed. A null object represents an entity or concept for which the fact repository engine 106 has no factual information and, as far as the fact repository engine 106 is concerned, does not exist. In some embodiments, a null object may be left in the fact repository 114. However, the null object is treated as if it was removed from the fact repository 114. In some other embodiments, null objects are removed from the fact repository 114.

FIG. 3 is a flow diagram illustrating a process for dereferencing attribute-value pairs, according to some embodiments of the invention. In a fact repository that is organized based on objects and facts (represented by attribute-value pairs, and optionally additional parameters as well) associated with objects, a fact may refer to other objects by name and/or by an identifier, such as an object ID. A fact that references another object may include only the object ID, or the object ID and an incorrect object name. The dereferencing process fills in the proper object name into the fact. The dereferencing process is similar to the functionality of the dereference operator in the C++ programming language. In C++, the dereference operator takes a pointer to a value and returns the value. In the process of FIG. 3, the fact repository engine takes a link to an object (the pointer) and inserts the name of the object.

A set of objects, stored in the fact repository 114, is accessed (302). The set of objects accessed may be the entirety of objects that are stored in the fact repository 114, or the set of accessed objects may be a subset of the entirety of objects stored in the fact repository 114. At least some of the accessed objects include one or more facts. In some embodiments, a fact, as described above in relation to FIG. 2, corresponds to an attribute-value pair (hereinafter called “A-V pair” for convenience). A fact may optionally be represented by additional parameters as well. It should be appreciated that while the description of FIGS. 3 and 4 below refer to A-V pairs, the description may be extended to facts having other, different data storage formats. Some of the A-V pairs may include links to other objects. That is, some of the A-V pairs may include links to objects other than the object with which the respective A-V pair is associated. The linked-to object may be any object in the fact repository 114 (including null objects) other than the object with which the A-V pair having the link is associated. The links may be the identifiers of the linked-to objects, as described above, in relation to FIG. 2. Some A-V pairs having links to objects may also include the names of the respective linked-to object in the value of the A-V pair, in addition to having the identifier of the linked-to object.

One of the objects in the set is selected (304). If the object does not include an A-V pair that includes a link to another object (306—no), nothing is done to that object. If the object includes one or more A-V pairs that include links to other objects (306—yes), the name of the respective linked-to object is inserted into the respective value of each A-V pair with a link to the linked-to object (308). If the value of an A-V pair already has a name for the linked-to object, in some embodiments that name is replaced by the inserted name, regardless of whether the pre-existing name is the same as the inserted name. In some other embodiments, the pre-existing name in an A-V pair is first compared against the name of the linked-to object. The name of the linked-to object in the A-V pair is replaced with the name of the linked-to object if the names do not match. In embodiments where a fact stores a link in the value field, the link is not replaced by the name when the name is inserted. Rather, the name and the link are concatenated and the concatenated string is stored in the value field.

If there are objects in the set remaining to be selected (310—no), another object is selected (304). Otherwise (310—yes), the process ends (312). The process may be repeated at scheduled intervals, or as needed.

In some embodiments, after operation 302, an optional table of object identifiers and object names may be built. The table maps object identifiers to their corresponding object names (in some embodiments, the primary names). The table may be loaded into memory. When inserting names into values, as described above, the fact repository engine 106 may refer to the table rather than searching for the object identifier in the fact repository itself. This may help make the dereferencing process more efficient.

As described above, an object may have more than one name. If the linked-to object has more than one name and one of them is designated the primary name, then the primary name is the one that is inserted into the value.

FIG. 4 is a flow diagram illustrating a process for checking references in attribute-value pairs, according to some embodiments of the invention. A set of objects, stored in the fact repository 114, is accessed (402). In some embodiments, the set of objects accessed is the entirety of objects that are stored in the fact repository 114. At least some of the accessed objects include one or more facts as A-V pairs. Some of the A-V pairs may include links to other objects in the set of objects.

One or more filters are applied to the A-V pairs of the objects and A-V pairs meeting predefined criteria are removed (404). The filters identify A-V pairs that meet predefined criteria and remove them. The predefined criteria may be defined based on the information conveyed by the A-V pair. For example, one predefined criterion for removal may be that an A-V pair is to be removed if it conveys a fact associated with pornography. In some embodiments, the predefined criteria may be implemented using heuristics and/or blacklists. The filters would apply the heuristics to the A-V pairs or compare the A-V pairs against a blacklist to determine which A-V pairs warrant removal. After the filtering, some objects may become null objects due to the removal of A-V pairs from the object.

One of the objects in the set is selected (406). If the object does not include an A-V pair that includes a link to another object or if all links in the A-V pairs of the object are to non-null objects (408—no), nothing is done to that object. If the object includes one or more A-V pairs that include links to null objects (408—yes), the links to null objects are removed (410).

In some embodiments, the removal of a link to a null object is performed by removing the identifier of the null object from the value of the A-V pair having the link to the null object. This link removal method is used only if there is already a name of the null object in the value of the A-V pair. In some other embodiments, the link is removed by removing the A-V pair (i.e., removing the fact) from the object. In further other embodiments, both manners of removal may be used; which one is used may depend on the circumstances with regard to how the linked-to object became a null object. For example, if the linked-to object became a null object because its associated A-V pairs were removed due to their meeting a first predefined criterion, then the manner of removal to be used may be removal of the A-V pair. On the other hand, if the associated A-V pairs were removed due to their meeting a second predefined criterion, then the manner of removal to be used may be removal of the identifier of the null object, leaving the name of the linked-to object in the value of the A-V pair.

If there are objects in the set remaining to be selected (412—no), another object is selected (406). Otherwise (412—yes), the process ends (414). In the embodiments where a manner of removal of the links to null objects includes removing the A-V pair with the link, instead of ending at 414, operations 406-412 may be repeated for additional iterations because the removal of the A-V pairs may create new null objects. While operations 406-412 may be repeated indefinitely, in some embodiments, a predefined limit on the number of additional iterations may be set. For example, after the first iteration, a limit may be set such that only one additional iteration is performed. It should be appreciated, however, that the process as illustrated in FIG. 4 may be performed at scheduled intervals or as needed. Furthermore, it should be appreciated that the filtering operation 404 and the reference checking operations 406-414 may be performed in accordance with different, independent schedules. For example, the filtering may be performed once a week and the reference checking may be performed once every 2 weeks. More generally, operations 406-414 do not have to be performed immediately after the filtering operation 404.

FIG. 5 is a block diagram illustrating a consistency maintenance system 500, according to some embodiments of the invention. The system 500 typically includes one or more processing units (CPU's) 502, one or more network or other communications interfaces 510, memory 512, and one or more communication buses 514 for interconnecting these components. The system 500 optionally may include a user interface 504 comprising a display device 506, keyboard 508 and pointer device 509, such as a mouse, track ball or touch sensitive pad. Memory 512 includes high-speed random access memory, such as DRAM, SRAM, DDR RAM or other random access solid state memory devices; and may include non-volatile memory, such as one or more magnetic disk storage devices, optical disk storage devices, flash memory devices, or other non-volatile solid state storage devices. Memory 512 may optionally include one or more storage devices remotely located from the CPU(s) 502. In some embodiments, the memory 512 stores the following programs, modules and data structures, or a subset thereof:

• • an operating system 516 that includes procedures for handling various basic system services and for performing hardware dependent tasks;
  • a network communication module (or instructions) 518 that is used for connecting the consistency maintenance system 500 to other computers via the one or more communication network interfaces 510 (wired or wireless), such as the Internet, other wide area networks, local area networks, metropolitan area networks, and so on;
  • a fact storage interface (or instructions) 520 that provides an interface to the fact storage system 530 (which may include a fact index and fact repository, and/or other appropriate data structures);
  • an object access module (or instructions) 522 that is used to access objects and associated attribute-value pairs;
  • an A-V pair filter (or instructions) 524 that removes attribute-value pairs that meet predefined criteria from objects;
  • a link reference checking module (or instructions) 526 that checks attribute-value pairs for links to null objects and removes the links to null objects; and
  • a link dereferencing module (or instructions) 528 that checks attribute-value pairs for links to objects and inserts the object names into the values of attribute-value pairs having links to objects.

The system 500 also includes a fact storage system 530 for storing facts. As described above, in some embodiments each fact stored in the fact storage system 530 includes a corresponding list of sources from which the respective fact was extracted.

Each of the above identified elements may be stored in one or more of the previously mentioned memory devices, and corresponds to a set of instructions for performing a function described above. The above identified modules or programs (i.e., sets of instructions) need not be implemented as separate software programs, procedures or modules, and thus various subsets of these modules may be combined or otherwise re-arranged in various embodiments. In some embodiments, memory 512 may store a subset of the modules and data structures identified above. Furthermore, memory 512 may store additional modules and data structures not described above.

Although FIG. 5 shows a “consistency maintenance system,” FIG. 5 is intended more as functional description of the various features which may be present in a set of servers than as a structural schematic of the embodiments described herein. In practice, and as recognized by those of ordinary skill in the art, items shown separately could be combined and some items could be separated. For example, some items shown separately in FIG. 5 could be implemented on single servers and single items could be implemented by one or more servers. The actual number of servers used to implement a consistency maintenance system and how features are allocated among them will vary from one implementation to another, and may depend in part on the amount of data traffic that the system must handle during peak usage periods as well as during average usage periods, and may further depend on the size of the fact repository and the amount of fact information each server can efficiently handle.

The foregoing description, for purpose of explanation, has been described with reference to specific embodiments. However, the illustrative discussions above are not intended to be exhaustive or to limit the invention to the precise forms disclosed. Many modifications and variations are possible in view of the above teachings. The embodiments were chosen and described in order to best explain the principles of the invention and its practical applications, to thereby enable others skilled in the art to best utilize the invention and various embodiments with various modifications as are suited to the particular use contemplated.