FORM 2
THE PATENTS ACT, 1970
(39 of 1970)
&
THE PATENTS RULES, 2003
COMPLETE SPECIFICATION (See section 10, rule 13)
1. Title of the invention: ENTERPRISE INFORMATION FUSION
2. Applicant(s)
NAME NATIONALITY ADDRESS
TATA CONSULTANCY Indian Nirmal Building, 9th Floor, Nariman Point,
SERVICES LIMITED Mumbai MH 400021
India
3. Preamble to the description
COMPLETE SPECIFICATION
The following specification particularly describes the invention and the manner in which it
is to be performed.

ENTERPRISE INFORMATION FUSION
TECHNICAL FIELD
[0001] The present subject matter, in general, relates to artificial intelligence, in
particular, to a system and a method for enterprise information fusion.
BACKGROUND
[0002] In presence of numerous news agencies and diversified factors that may affect
business operation of an enterprise, it is a challenge for business analysts to monitor and track
news reported by the new agencies and online content posted by general public on the World
Wide Web. Online content may include customers feedback posted on various blogs, forums,
websites, etc. related to various products/services provided by the enterprise.
[0003] In general, the news may include news related to various events, such as
incidents or happenings occurring at international, state or local level, and customer feedback related to various products/services provided by the enterprise. Some of the events may impact the business of the enterprise. For example, earthquake in Japan is an event that may have an adverse impact on the business of the enterprise having factories of their key suppliers in Japan. In another example, problems faced by customers of a XYZ product of an enterprise, may affect the business of the enterprise. Accordingly, it is required for the enterprise to be aware of events that impact the business of the enterprise. For awareness, business analysts within the enterprise read the news from a plurality of sources and also subscribe to suitable newsletters, RSS feeds from various blogs, forums, etc. which report on specific events, to identify various events from the news that may have an impact on the business of the enterprise. Upon identifying such events, the business analysts analyzes such events and manually correlate these events with information associated with the operations of the enterprise to asses risk associated with such new events on the business operations.
SUMMARY
[0004] This summary is provided to introduce concepts related to enterprise
information fusion. These concepts are further described below in the detailed description.

This summary is not intended to identify essential features of the claimed subject matter nor is
it intended for use in determining or limiting the scope of the claimed subject matter.
[0005] Systems and methods related to enterprise information fusion are described.
The method comprises obtaining event information corresponding to at least one event, wherein the at least one event includes at least one of an incident and a customer feedback. Based on the event information, information corresponding to at least one entity associated with at least one operation of an enterprise is retrieved, from a plurality of entity information sources. An impact of the at least one event on the at least one entity is determined. Based on the determination of the impact, a risk associated with the at least one event on the at least one operation of the enterprise is evaluated. An alert is generated, based on the evaluation, where the alert is indicative of the risk.
BRIEF DESCRIPTION OF THE DRAWINGS
[0006] The detailed description is provided with reference to the accompanying
figures. In the figures, the left-most digit(s) of a reference number identifies the figure in
which the reference number first appears. The same numbers are used throughout the
drawings to reference like features and components.
[0007] Fig. 1 illustrates a network environment implementing an enterprise
information fusion system, in accordance with an embodiment of the present subject matter.
[0008] Fig. 2 illustrates components of the enterprise information fusion system, in
accordance with an implementation of the present subject matter.
[0009] Fig. 3 illustrates an exemplary method for enterprise information fusion
system, in accordance with an implementation of the present subject matter.
DETAILED DESCRIPTION
[00010] Conventionally, business analysts within an enterprise read the news from a
plurality of sources and also subscribe to suitable newsletters, RSS feeds from various blogs, forums etc. which report on specific events, in order to identify the events that may have adverse impact on the business of the enterprises. Upon identifying such events, the business analyst manually analyzes and correlates such events with information associated with operations of the enterprise to study and evaluate impact of these events on the business of the

enterprise. Such conventional approach of event detection is ineffective as these approaches have greater possibility of missing out on some events. Generally, some events, such as incidents that are of local importance only are not reported in the mainstream news, consequently such news get missed out by the business analysts inevitably. For example, fire in a factory of supplier's key suppliers is an event that may not be reported in the mainstream news and therefore is likely to be missed out by the business analysts. The event indicated by above example is news that may have adverse impact on the business of the enterprise. Further, event such as customer feedback that is typically present in unstructured format and scattered across the web in various blogs, forums, etc. is likely to be missed from the business analysts in part or completely. Therefore, missing out events may keep the enterprise unaware of risk associated with such events on the enterprise operations.
[00011] In case the business analyst identifies the events that could be of relevance to
an enterprise, from the news, such conventional approaches require the business analyst to
manually analyze and correlate these events with information associated with the operation of
the enterprise, such as supply chain information, information pertaining to various contracts,
sales information etc. to asses risk associated with such events on the business of the
enterprise. Such manual analysis to asses the risk associated with such events requires the
business analyst to have in-depth knowledge and details of the business operation. For
example, if an event indicates that there is fire in a factory of supplier's key suppliers, then
business analyst should be aware of the details of all the supplier’s key suppliers in order to
identify and correlate such event with the business operation, to study the impact of the event
on the operations of the enterprise and assess risk associated therewith, which is practically
not feasible. Further, manual analysis and correlation process requires lot of efforts and time.
Furthermore, manual analysis is error prone as even skilled and experienced business analysts
at times miss-out on most of the factors which affect the enterprise operations.
[00012] To this end, systems and methods for enterprise information fusion are
described. The system monitors news from a plurality of news sources and extracts event information pertaining to one or more events from the news. The events may include one or more of an incident and a customer feedback pertaining to various products and services of an enterprise. The system correlates the event information pertaining to the events with information about the enterprise to evaluate an impact of each of the one or more events on

the operations of the enterprise. Information may be internal or external to the enterprise. If the result of the evaluation indicates a threat to one or more operations of the enterprise, the system generates real-time alerts indicative of the impact of the events on the one or more operations of the enterprise.
[00013] In an implementation, the system mines the news from a plurality of news
sources. The plurality of news sources may include external sources, namely, the World Wide
Web (web) and internal sources, i.e., information available within the enterprise. The news
may include information pertaining to various events, such as incidents occurring at an
international, state or local level, and customer feedback pertaining to various products and
service provided by the enterprise, and other general news. Based on the mining, the system
extracts event information pertaining to various events, such as incidents and customer
feedback. The event information may indicate, for example, event details indicating what has
happened, place of occurrence of the event, time at which the event has occurred etc.
[00014] In an implementation, the system extracts incident information pertaining to
various incidents from social media, such as Twitter™. As a consequence of the growth and
popularity of social media such as Twitter™, incidents of even minor or highly local
importance is often reported in Twitter™ by reporters as well as the general public. Therefore,
there is reduced chance of missing out any incident that could be of relevance to the
enterprise. In said implementation, the system extracts customer feedback from various blogs,
forums, websites, etc. Additionally, the system extracts customer feedback from data
available within the enterprise, such as direct email communications containing consumer
complaints and suggestions as well as transaction logs generated as a result of direct customer
communications. The system therefore mines both external sources and internal sources to
detect a plurality of events, i.e., incidents and customer feedback. The system may also detect
change in sentiments of the customer by continually monitoring the customer’s feedback.
[00015] Subsequent to detecting the event information pertaining to the plurality of the
events, the system may gather additional information pertaining to the events from the web. The system thereafter, stores the event information in an event data repository associated with the system.
[00016] Based on the event information, the system retrieves various enterprise facts
from a plurality of entity information sources. The enterprise facts are indicative of the

information about one or more entities associated with one or more operations of an enterprise. Such entities may include various suppliers, vendors, distributors, and customers associated with the one or more operations of the enterprise. Examples of various enterprise facts include name of the entities associated with the enterprise either directly or indirectly, their relationships with the enterprise, and geographical location of such entities. The plurality of entity information sources referred herein includes internal information sources, i.e., data sources within the enterprise, and external information sources, such as open source information available on the World Wide Web (web). The open source information may be understood as publically available and freely accessible information on the web. In one implementation, the enterprise facts from the open source information available on the web are retrieved in real-time, subsequent to extraction of the event information. In said implementation, the web is, therefore, used as one of the entity information sources. In another implementation, the enterprise facts from the web are retrieved prior to the extraction of event information and stored in an enterprise facts repository associated with the system. In said implementation, the enterprise facts repository is used as one of the entity information sources, from where one or more enterprise facts are retrieved subsequent to extraction of the event information. The system thereafter evaluates the impact of each of the events on the one or more entities associated directly or indirectly with the one or more operations of the enterprise by fusing the event information pertaining to the event with the enterprise facts. Fusion of the event information with the enterprise facts may be understood as correlating the event information with the enterprise facts to analyze the impact of the event on the one or more entities associated with the enterprise. In an implementation, the system refers to a plurality of pre-defined rules for analyzing the impact.
[00017] Based on the result of the analysis, the system further evaluates if the impact
on an entity associated with the enterprise represents a threat or risk to the operations of the enterprise. For evaluation, the system retrieves additional facts about the entity under impact from the enterprise data. The additional facts referred herein are indicative of the various operational and/or transactional details associated with the entity. The operational and/or transactional details may indicate various contracts, shipments, assignments, sales etc. corresponding to the entity. For example, if an impact on a supplier associated with an enterprise is identified, and the operational and/or transactional details associated with the

supplier indicates that there are supplies pending from the supplier, such an impact on the supplier indicate a risk to the supply chain activities of the enterprise. Based on the result of the risk evaluation, the system generates an alert. For example, if the result of the evaluation indicates a risk to the enterprise operation, the system generates an alert indicative of the risk that may occur due to the event. The enterprise may therefore take appropriate actions to prevent or deal with the risk. In an implementation, the alert further indicates probability of the risk in terms of percentage.
[00018] The system in accordance with the present subject matter automatically
extracts and analyzes the event information corresponding to the one or more events and correlates the event information with the information associated with the enterprise to evaluate the impact of the one or more event on the enterprise. Therefore, the system eliminates the manual work required by the business analysts to read the news from the plurality of sources, identify various events from the news, analyzes the impact of the events on the one or more operation of the enterprise by studying and correlating the event information corresponding to the events with the information associated with the one or more operations of the enterprise. Further, the system facilitates mining of large number of events and evaluates impact associated with them. Moreover, the system also reduces the possibility of errors or missing out events that may have an impact on the enterprise. Further, the system is capable of generating real-time alerts on detecting any event which poses a potential risk to the enterprise operations. Therefore, the enterprise may become aware of the risk associated with the event on the enterprise operations, at an early stage, so that the enterprise can take appropriate risk management actions at a right time, thereby preventing/reducing the risk associated with such event.
[00019] While aspects of systems and methods for enterprise information fusion can be
implemented in any number of different computing systems, environments, and/or configurations, the embodiments are described in the context of the following exemplary system architecture(s).
EXEMPLARY SYSTEMS
[00020] Fig. 1 illustrates an exemplary network environment 100 implementing an
enterprise information fusion system 102, in accordance with an embodiment of the present

subject matter. In one implementation, the network environment 100 can be a company network, including thousands of office personal computers, laptops, various servers, such as blade servers, and other computing devices connected over a network 106. In another implementation, the network environment 100 can be a home network with a limited number of personal computers and laptops connected over the network 106.
[00021] The enterprise information fusion system 102 (hereinafter referred as system
102) is connected to a plurality of user devices 104-1, 104-2, 104-3,...104-N, collectively referred to as the user devices 104 and individually referred to as a user device 104. The system 102 and the user devices 104 may be implemented as any of a variety of conventional computing devices, including, servers, a desktop personal computer, a notebook or portable computer, a workstation, a mainframe computer, a mobile computing device, and a laptop. The system 102 is connected to the user devices 104 over the network 106 through one or more communication links. The communication links between the system 102 and the user devices 104 are enabled through a desired form of communication, for example, via dial-up modem connections, cable links, digital subscriber lines (DSL), wireless or satellite links, or any other suitable form of communication.
[00022] The network 106 may be a wireless network, a wired network, or a
combination thereof. The network 106 can also be an individual network or a collection of many such individual networks, interconnected with each other and functioning as a single large network, e.g., the Internet or an intranet. The network 106 can be implemented as one of the different types of networks, such as intranet, local area network (LAN), wide area network (WAN), the internet, and such. The network 106 may either be a dedicated network or a shared network, which represents an association of the different types of networks that use a variety of protocols, for example, Hypertext Transfer Protocol (HTTP), Transmission Control Protocol/Internet Protocol (TCP/IP), etc., to communicate with each other. Further, the network 106 may include network devices, such as network switches, hubs, routers, for providing a link between the system 102 and the user devices 104. The network devices within the network 106 may interact with the system 102 and the user devices 104 through the communication links. The users such as business analysts may use the user devices 104 to interact with the system 102 and to view the alerts generated by the system 102.

[00023] The system 102, according to an implementation of the present subject matter
extracts event information pertaining to one or more events from the news. The events may include incidents occurring at an internal, state, or local level and customer feedback pertaining to various products and service of an enterprise. The event information may indicate, for example, place of occurrence of the event, time at which the event has occurred, and event details indicating what has happened.
[00024] Once the event information pertaining to the one or more events is extracted,
the system 102 may gather additional details pertaining to such events from the web. The system 102 stores the extracted event information within an event data repository associated with the system 102. Based on the event information, the system 102 retrieves various enterprise facts from a plurality of entity information sources. The plurality of entity information sources referred herein includes internal information sources, i.e., data sources within the enterprise, and external information sources, such as open source information available on the World Wide Web (web). In one implementation, the enterprise facts from the external information sources, such as the web are retrieved in real-time subsequent to extraction of the event information. The web is therefore used as one of the entity information sources in said implementation. In another implementation, the enterprise facts from the web are retrieved before the extraction of event information and stored in an enterprise facts repository associated with the system, where the enterprise facts repository is used as one of the entity information sources. In said implementation, upon extraction of the event information, one or more enterprise facts are retrieved from the enterprise facts repository, based on the event information. The enterprise facts are indicative of the information pertaining to one or more entities, such as various vendors, suppliers, distributors, customers, etc., associated with the one or more operations of the enterprise. Based on the enterprise facts, and the event information, the system 102 determines an impact of each of the events on the one or more entities.
[00025] Based on the result of the analysis, the system 102 evaluates the risk associated
with the impact on the operations of the enterprise. For evaluation, the system 102 retrieves enterprise data associated with the one or more entities under the impact. The enterprise data is indicative of operational and/or transactional details associated with the entity under impact. The system 102 thereafter analyzes the retrieved enterprise data and identifies any

risk on the one or more operations of the enterprise, such as supply chain operations, sales
operations, enterprise performance management, and customer relationship management.
[00026] In an implementation, the system 102 comprises a risk evaluation module 108
for evaluating the risk associated with the event on the operations of the enterprise. For evaluation, the risk evaluation module 108 retrieves various enterprise facts from the plurality of entity information sources, based on the event information associated with the event under risk evaluation. The enterprise facts as referred herein are indicative of information pertaining to the one or more entities associated with the one or more operations of the enterprise. Such entities may either have a direct association or an indirect association with the one or more operations of the enterprise. In case of direct associations, an entity and the enterprise is directly linked to one another without the involvement of a third party in between. While in case of indirect associations, an entity and the enterprise is linked to one another with one or more entities in between. For example, an entity A may have direct association with an entity B, and the entity B has direct association with the enterprise. In said example, the entity A and the enterprise have indirect association with one another.
[00027] The enterprise may have information or facts about various entities directly
associated with the enterprise. However, information about the entities indirectly associated with the enterprise may not be available within the enterprise. Therefore, such information is retrieved from the web, based on the event information. For example, if the event information ‘industrial unrest in Pune’ is identified pertaining to an event ‘industrial unrest’ extracted, the enterprise may have the name and geographical location information related to some entities directly associated with the enterprise. However, name and geographical location information related to entities that are indirectly associated with the enterprise may not be available with the enterprise. In said example, there may be a case where one or more entities that are indirectly associated with the enterprise have their factories in pune, and industrial unrest in pune may have an impact on such entities, which in-turn may have an impact on the entities directly associated with the enterprise. In operation, the risk evaluation module 108 retrieves such information (also referred as enterprise facts) from the World Wide Web (web). In an implementation, the risk evaluation module 108 retrieves such information from the web in real-time, based on the extracted event information. In another implementation, the risk evaluation module 108 retrieves such information from the information pre-stored in the

enterprise facts repository associated with the system. For retrieving such information from the web or the enterprise facts repository, the risk evaluation module 108 performs similarity search that identify both the exact information and related information based on the keywords indicated in the event information. In the example mentioned above, the risk evaluation module 108 retrieves data corresponding to the one or more entities, such as suppliers, vendors, distributors etc. located in and around Pune, as Pune has been indicated as the place of occurrence of the event in the event information. Assuming that the search result retrieves a fact ‘XYZ motor has a factory in Pimpri’, which is a place around Pune, the risk evaluation module 108 stores this information or fact as the enterprise facts within the enterprise facts repository associated with the system 102.
[00028] The risk evaluation module 108 further retrieves information, about the one or
more entities, available within the enterprise from the enterprise data. For retrieval, the risk evaluation module 108 conducts a search within the enterprise data to retrieve facts related to the entities directly associated with the enterprise. The risk evaluation module 108 stores such facts as the enterprise facts within the enterprise facts repository. In an implementation, the risk evaluation module 108 evaluates the entity impact, i.e., impact of the event on various entities, such as supplier and vendors associated with the one or more operations of the enterprise. Impact on any entity associated with the enterprise, may in turn impact the one or more operations of the enterprise. For example, if an event indicates that there is industrial unrest in Pune, and the enterprise may have one of their key-suppliers, which in-turn is supplied by a supplier located in or around Pune, the supply chain operations of an enterprise may be affected. The risk evaluation module 108 therefore retrieves data from enterprise data about various entities and their relationship with the enterprise, based on the entities indicated by the facts retrieved from the web and/or event information.
[00029] In the example mentioned above, on retrieving the facts ‘XYZ motor having
factory in Pimpri (near Pune)’ from the web, the risk evaluation module 108 conducts a search within the enterprise data to locate additional facts about the entity XYZ motors. If the result of the search indicates that ‘ABC is a supplier of the enterprise’, and ‘XYZ motors is a supplier of ABC’, the risk evaluation module 108 analyses the retrieved facts that indicates that ABC, one of the suppliers of the enterprise, is supplied by the XYZ motors, and the XYZ motors is located around Pune. Further, the event information pertaining to the event under

impact evaluation indicates that there is industrial unrest in ‘Pune’. The risk evaluation module 108 therefore correlates these facts with the event information. In the same example mentioned above, the risk evaluation module 108 may evaluate that the XYZ motors could be impacted by the event industrial unrest, which in turn may have an impact on the ABC, and the enterprise.
[00030] Further, the risk evaluation module 108 evaluates the risk associated with the
impact on the one or more operations of the enterprise. For evaluation, the risk evaluation module 108 retrieves enterprise data indicative of the operational and/or transactional details association with the entity under impact. Examples of the enterprise data may include various contracts and shipments pending from the supplier end, transactions data and sales data. In an example, if a supplier impact is identified, and the enterprise data indicates that the supply from impacted supplier is pending, such an impact is indicative of a potential threat or risk to the supply chain operations of the enterprise. The risk evaluation module 108 identifies the risk associated with such impacts on the one or more operations of the enterprise. In an implementation, the risk evaluation module 108 also determines the probability of the risk in form of percentage. The risk evaluation module 108 further generates alerts, based on the result of risk evaluation. The alerts may indicate the risk indicated by the event on the one or more operations of the enterprise, thereby making the enterprise aware of such risk. The enterprise may therefore take appropriate actions to prevent or deal with the risk associated with the event on the operations of the enterprise. The generated alerts may be communicated to the enterprise in form of email communication, Short Message Service (SMS), postings on a dashboard etc.
[00031] Fig. 2 illustrates exemplary components of the enterprise information fusion
system 102 for evaluating risk associated with various events, such as incidents and customers feedback on the one or more operations of the enterprise, according to an embodiment of the present subject matter. In said embodiment, the enterprise information fusion system 102 includes one or more processor(s) 202, a memory 204 coupled to the processor 202, and interface(s) 206. The processor 202 may be implemented as one or more microprocessors, microcomputers, microcontrollers, digital signal processors, central processing units, state machines, logic circuitries, and/or any devices that manipulate signals based on operational

instructions. Among other capabilities, the processor 202 is configured to fetch and execute computer-readable instructions and data stored in the memory 204.
[00032] The interfaces 206 may include a variety of software and hardware interfaces,
for example, interface for peripheral device(s) such as a keyboard, a mouse, an external memory, a printer, etc. Further, the interfaces 206 may enable the enterprise information fusion system 102 to communicate with other computing devices, such as web servers and external databases. The interfaces 206 may facilitate multiple communications within a wide variety of protocols and networks, such as a network, including wired networks, e.g., LAN, cable, etc., and wireless networks, e.g., WLAN, cellular, satellite, etc. The interfaces 206 may include one or more ports for connecting the enterprise information fusion system 102 to a number of computing devices.
[00033] The memory 204 may include any computer-readable medium known in the art
including, for example, volatile memory such as static random access memory (SRAM) and
dynamic random access memory (DRAM), and/or non-volatile memory, such as read only
memory (ROM), erasable programmable ROM, flash memories, hard disks, optical disks, and
magnetic tapes. The memory 204 also includes module(s) 208 and data 210.
[00034] The modules 208 include routines, programs, objects, components, data
structures, etc., which perform particular tasks or implement particular abstract data types. The modules 208 further include an event extraction module 212, a risk evaluation module 108, and other modules 214. The other modules 214 may include programs or coded instructions that supplement applications and functions on the enterprise information fusion system 102, for example, programs in the operating system.
[00035] The data 210, amongst other things, serves as a repository for storing data
processed, received, and generated by one or more of the module(s) 208. The data 210 includes event data 216, enterprise facts 218, enterprise data 220, evaluation data 222, rules 224, alerts 226, and other data 228. The other data 228 may include data generated as a result of the execution of one or more modules in the other modules 214.
[00036] In operation, the enterprise information fusion system 102 (hereinafter referred
as system 102) monitors news from a plurality of sources, extracts event information corresponding to one or more events such as incidents and customers feedback from news, and fuses the event information with information from within and about the enterprise to

evaluate the impact of the one or more events and generate real-time alerts based on the evaluation. The entire procedure of extracting event information, and fusion of the event information with the information within and about the enterprise to evaluate the risk associated with the events on the one or more operations of the enterprise and generation of alerts is explained in detail under the following sections, viz., event information extraction, risk evaluation, and alerts generation.
Event Information Extraction
[00037] In operation, the event extraction module 212 extracts event information
corresponding to one or more events, such as incidents and customers feedback from the news. The manner in which extraction of the event information is performed is explained in detail under the following sub-sections, viz., incident information extraction and customer feedback analysis.
Incident Information Extraction
[00038] The event extraction module 212 extracts incident information corresponding
to one or more incidents from external sources, such as World Wide Web (web). In an implementation, the event extraction module 212 extracts incident information from social media, such as Twitter™. The social media, such as Twitter™ contains news, updates, personal communications posted by a variety of users, such as news reporters, celebrities, business professionals, students, general public etc., in form of short messages. Such social media in general contain profiles of various users. The user profiles allow users to post messages/read messages in/from their own profiles or profiles of other users, if allowed by other users. The user profiles also offer users to perform various additional activities, apart from posting/reading messages, depending upon the type of social media. In social media, such as Twitter™, such short messages are referred as tweets. While, in social media, such as Facebook™, these short messages are referred as posts. Further, in social media, such as Orkut™ these short messages are referred as scraps. The short messages may include news about various incidents or happening around the world, views or thoughts of various users about such incidents or happenings, personal communications messages or updates etc. With the growth and popularity of the social media, news of even minor or highly local importance

is often reported here by news reporters as well as the general public in the form of short messages.
[00039] In said implementation, the event extraction module 212 identifies a set of user
profiles, for example, the profiles of the users known for reporting news on various incidents, community profiles or profiles created by various users, and other users listed in a friend or follower list. Subsequent to identification, the event extraction module 212 filters the set of user profiles, based on pre-defined keywords, i.e., the event extraction module 212 performs a search on profile information (also referred as bibliographic profile information) based on certain pre-defined keyword, to create a filtered set of user profiles. The extracting module 212 continually monitors the filtered set of user profiles for obtaining the short messages posted on such profiles. The short messages may be in form of tweets, posts, scraps etc. based on the social media from where such short messages are retrieved. These short messages may be indicative of the occurrence of various incidents. The event extraction module 212 extracts incident information from the short messages, based on various pre-defined rules stored within rules 224.
[00040] In an implementation, the event extraction module 212 stores all the incident
information corresponding to same incident in a buffer. In other words, a series of short messages may relate to same incident. In such a case, the incident information pertaining to the same incident is stored in a buffer, such information is then collated and stored a repository event data 216. The incident information may include, for example, place of occurrence of the incidents, time of occurrence of the incidents, details of the incident indicating what has happened, etc. Further, some short messages indicating about an incident may carry a link to a published material that potentially contains detailed information on the reported incident. In such cases, the event extraction module 212 performs web-crawling, i.e., follow the link and extract additional incident information about the incident therefrom, in order to model the incident in detail. The event extraction module 212 stores such incident information in the event data 216.
[00041] The incident extraction is explained with reference to social media only for the
purpose of explanation and not as a limitation. It will be appreciated by the person skilled in the art that incidents may also be detected from other sources on the web, such as various news websites.

Customer Feedback Analysis
[00042] In addition to the incident information corresponding to the plurality of
incidents, the event extraction module 212 extracts customers feedback from external sources,
such as World Wide Web (web) and internal sources, such as enterprise data. It is to be
understood that the enterprise may receive customer communications, such as customer
feedback regarding the products and/or services provided by the enterprise, over email, call-
center conversations, which may be transcribed to text and stored within the enterprise.
Further, the customers may post responses to surveys or questions posted on the website of
the enterprise, which are collated within a repository within the enterprise. Such email
communications, call-center conversations, customers posted response on enterprise websites
forms part of the enterprise data, which is also referred to as internal data sources.
[00043] Further, the customers may post comments related to services and products of
the enterprise, on various blogs, discussion forums etc. Such blogs, forums etc. are referred to as external data sources. The customer response, comments, communications etc. are collectively referred to as customer feedback. It is to be noted that the customer feedback is generally present in an unstructured format. The event extraction module 212 retrieves these customers feedback from the various sources, i.e., external data sources and internal data sources, using various text mining techniques, known in the art. Subsequent to retrieval, the event extraction module 212 uses conventionally known Natural Language Processing (NLP) techniques, to automatically analyze the customer feedback therefrom. Further, the event extraction module 212 may extracts details of the customer feedback. The details includes, for example, type of product/service related to which feedback is provided by the customers, nature of the feedback, such as positive, negative, neutral etc. The event extraction module 212 extracts the details in structured format and stores the extracted details in the event data 216. In an implementation, the event extraction module 212 may group the customers feedback based on various parameters, such as type of product/services, nature of the feedback, etc.
[00044] For the sake of clarity, the incident and the sentiments are collectively referred
to as events, and the incident information corresponding to the one or more incidents and information related to the customer feedback is collectively referred to as event information.

Risk Evaluation
[00045] Once the event information corresponding to the one or more events, i.e.,
incidents and customers feedback are extracted from the news, the risk evaluation module 108 evaluates the impact of the each of the one or more events on the one or more operations of the enterprise.
[00046] In an implementation, the risk evaluation module 108 uses a Blackboard
architecture reasoning technique, known in the art to evaluate the impact of the events. Event
information corresponding to a large number of events is detected rapidly by the event
extraction module 212. Evaluating the impact of an event may involve exploring many
different possibilities based on facts about the enterprises business. Such facts are usually
needed to be extracted from the enterprises information systems. Further, multiple events,
occurring at different times, may collectively contribute to potential impact. Therefore, the
risk evaluation module 108 uses a Blackboard architecture reasoning technique, known in the
art that continuously process event information on a blackboard, correlate them with
enterprise facts, as well as extract new enterprise facts as needed from enterprise data. The
blackboard technique implements a parallel terraced scan, i.e., exploring multiple possibilities
simultaneously so that the chances of missing something of real importance are minimized.
[00047] In operation, if event information corresponding to an event is extracted and
placed on the blackboard. The risk evaluation module 108 evaluates the impact of the event on the operations of the enterprise. For evaluation, the risk evaluation module 108 retrieves various enterprise facts from a plurality of entity information sources, based on the event information. The enterprise facts are indicative of the information associated with one or more entities associated with the operations of the enterprise. The plurality of entity information sources for the retrieval of such enterprise facts include internal information sources, i.e., data sources within the enterprise, and external information sources, such as open source information available on the World Wide Web (web). In one implementation, the enterprise facts from the external information sources, such as the web are extracted in real-time from the web, when the event information corresponding to one or more events is extracted. In said implementation, web is used as one of the entity information sources. In another implementation, the enterprise facts from external information sources, such as the web are extracted and stored in an open source information repository associated with the system. In

said implementation, when the event information is extracted, one or more enterprise facts based on the event information is retrieved from the repository, which is referred as one of the entity information sources. Subsequent to retrieving the enterprise facts, the risk evaluation module 108 analyzes and correlates the retrieved enterprise facts with the event information to determine the impact of the events on the one or more entities associated with the enterprise. If the result of the evaluation indicates an impact, the risk evaluation module 108 evaluates risk associated with the impact on the one or more operations of the enterprise. The procedure in which risk evaluation takes place is explained in detail under the following sub-sections, viz, entity impact evaluation and risk evaluation on the enterprise operations.
Entity Impact Evaluation
[00048] Based on the event information corresponding to the one or more events, i.e.,
incidents and customer feedback, the risk evaluation module 108 extracts information about various entities associated with the enterprise from the information available within the enterprise and the open source information available on the web. The enterprise may have different entities, such as suppliers, vendors, distributors etc. associated with the enterprise either directly or indirectly. The enterprise may have information about entities that are directly associated with the enterprise. However, it is unlikely that the enterprise may have information about entities that are indirectly associated with the enterprise. For example, the enterprise may have information about its key-suppliers. However, information about other suppliers that supply to the key-suppliers of the enterprise may not be available with the enterprise. As an instance, if XYZ motors is a supplier to the enterprise, the XYZ motors may in-turn be supplied by a supplier ABC motors. In said instance, it is unlikely that the enterprise is aware of the fact that ABC motor is a supplier to the XYZ motor. Impact on any of the direct or indirect entity associated with the one or more operations of the enterprise, may pose a risk on the one or more operations of the enterprise. With reference to the instance mentioned above, if an event indicating fire in a factory of ABC motor is identified, the supply chain activities of the enterprise may be affected. Therefore, information about such entities indirectly associated with the enterprise is required to be retrieved from the web. In operation, the risk evaluation module 108 retrieve information about the various entities associated with the enterprise from the web that may not be available within the enterprise but

may have an impact on the operations of the enterprise. The risk evaluation module 108 extracts such information, i.e., the enterprise facts from the web, based on the event information.
[00049] In an example, an event information ‘industrial unrest’ in place ‘Pune, India’ is
extracted. The risk evaluation module 108 searches the web for various enterprise facts related to the one or more entities associated with the one or more operations of the enterprise. For searching the enterprise facts on the web, the event extraction module 212 performs similarity search to identify both the exact information and related information based on the keywords indicated in the event information. In said example, the event extraction module 212 retrieves data corresponding to all the suppliers, vendors, distributors etc. located in Pune or around Pune. If the search results indicates that XYZ motors has a factory in Pimpri, which is around Pune which is indicated by the event information. The risk evaluation module 108 searches for additional facts about the XYZ motors to determine if XYZ motor is related to the enterprise through any relationship. Such determination is made by performing entity association search by referring to information about various entities stored within the enterprise, i.e., enterprise data 220, and performing additional search on the web. Assuming that the search result indicates that XYZ motor is a supplier to ABC motors, and ABC motor is a supplier to the enterprise, the risk evaluation module 108 stores these facts as enterprise facts within the enterprise facts 218.
[00050] The risk evaluation module 108 analyses and correlates these enterprise facts
with the event information to determine if the event has an impact on the enterprise. In the example mentioned above, the risk evaluation module 108 analyses and correlate the retrieved enterprise facts with the event information to determine that ABC is a supplier to the enterprise, which in-turn is supplied by XYZ motors having factory around Pune, where industrial unrest has happened. The risk evaluation module 108 therefore determines a supplier impact considering that the XYZ motors could be impacted by the event “industrial unrest”, which in turn may have an impact on the ABC, and impact on the ABC may impact the enterprise.
Risk Evaluation on Enterprise Operations

[00051] The risk evaluation module 108 further evaluate if the impact indicates a risk
on the one or more operations of the enterprise by further correlating the result of the impact with additional enterprise facts related to the one or more operations of the enterprise. The enterprise usually contains data related to the operations and transactions pertaining to different entities, i.e., vendors, suppliers, customers, etc. stored in various repositories within the enterprise, or scattered in from of documents, presentations, email communications, etc. within the enterprise. For large enterprises total size of this data may be in multiple terabytes or sometimes even in petabytes. Based on impact analysis on the entities associated with the enterprise, the risk evaluation module 108 extracts details related to enterprise operations stored within the enterprise. In another example, if the impact analysis indicates an impact on entity ABC motors that is one of the suppliers of the enterprise, the risk evaluation module 108 performs an object search with the enterprise data, based on the entity under impact, i.e., ABC motors. Therefore, the entity ABC motors may be considered as an object to be searched within the enterprise data. By performing the object search the risk evaluation module 108 extracts enterprise data 220 about the entity. The enterprise data 220, for example, may include the records pertaining to the entity, the supplier record for the supplier, shipments already dispatched from the supplier, procurement contracts placed on, and the products supplied by XYZ Company, etc. Based on the enterprise data, the risk evaluation module 108 evaluates the impact of the event on the operations of the enterprise. For example, if an impact on a supplier ABC motors is identified, and the enterprise data indicates that all supplies from the ABC motors has been received, the risk evaluation module 108 therefore indicates that there is no risk on the supply chain activities of the enterprise because of the event. In another example, if an impact on supplier ABC motor is identified and there are supplies pending from the ABC motor, the risk evaluation module 108 therefore indicates a risk on the supply chain operation of the enterprise.
Alert Generation
[00052] Based on the result of risk evaluation, the risk evaluation module 108 generates
alerts indicative of the risk associated with the event on the one or more operations of the

enterprise, thereby making the enterprise aware of the risk. The enterprise may therefore take appropriate actions to prevent or manage risk associated with the event on the operations of the enterprise. The risk evaluation module 108 communicates the generated alerts to the enterprise in form of email communication, SMS, postings on a dashboard etc. In an implementation, the alerts further indicate probability of the risk associated with the event in form of percentage. For example, ‘Potential threat to supplies due to earthquake in New-Zealand with probability 79.13 %’ may be generated as an alert, if an event indicates an impact on the suppliers located in New-Zealand, where supplies from the suppliers are pending.
[00053] Fig. 3 illustrates an exemplary method 300 for enterprise information fusion, in
accordance with an implementation of the present subject matter. The exemplary method may be described in the general context of computer executable instructions. Generally, computer executable instructions can include routines, programs, objects, components, data structures, procedures, modules, functions, etc., that perform particular functions or implement particular abstract data types. The method may also be practiced in a distributed computing environment where functions are performed by remote processing devices that are linked through a communications network. In a distributed computing environment, computer executable instructions may be located in both local and remote computer storage media, including memory storage devices.
[00054] The order in which the method is described is not intended to be construed as a
limitation, and any number of the described method blocks can be combined in any order to implement the method, or an alternative method. Additionally, individual blocks may be deleted from the method without departing from the spirit and scope of the subject matter described herein. Furthermore, the method can be implemented in any suitable hardware, software, firmware, or combination thereof.
[00055] At block 302, event information corresponding to at least one event is
extracted from the news. In one implementation, the event extraction module 212 of the enterprise information fusion system 102 monitors news from a plurality of sources The plurality of sources may include external sources, namely, the World Wide Web (web) and internal sources, i.e., information available within the enterprise. The news may include information pertaining to various events, such as various incidents/happenings occurring at

international, state or local level, and customer’s feedback related to various products and services of an enterprise, and other general news. The event extraction module 212 extracts event information corresponding to at least one event from the news. The event information may include, for example, a place of occurrence of the event, time of occurrence of the event, date of occurrence of the event, etc. The event may be an incident or a customer feedback. The event extraction module 212 store the extracted event information in an event data repository, such as the event data 216 associated with the enterprise information fusion system 102.
[00056] At block 304, a plurality of enterprise facts from information available within
the enterprise and outside the enterprise are retrieved, based on the event information. The enterprise may have different entities, such as suppliers, vendors, distributors etc. associated with the enterprise either directly or indirectly. The enterprise may have information about entities that are directly associated with the enterprise. However, the enterprise may have information about entities that are indirectly associated with the enterprise. Impact on any entity associated either directly or indirectly with the enterprise may in-turn have an impact on the operations of the enterprise. Therefore, information related to various entities associated with the enterprise, i.e., the enterprise facts are required to be retrieved from the information available within the enterprise, i.e., the enterprise data, such as enterprise data 220, and the information available outside the enterprise, such as the World Wide Web (web). In one embodiment, the risk evaluation module 108 within the enterprise information fusion system 102 retrieves the enterprise facts from the enterprise data 220 and the web. The risk evaluation module 108 stores these enterprise facts in an enterprise facts repository, such as enterprise facts 218 associated with the enterprise information fusion system 102. Examples of the enterprise facts may include name, address, geographical location of various entities associated with the enterprise.
[00057] At block 306, impact of the event on the one or more entities associated with
the enterprise is determined, based on the enterprise facts. In operation, the risk evaluation module 108 determines the impact of the event on the one or more entities associated with the enterprise, by correlating the event information with the enterprise facts. For example, if an enterprise fact indicates that XYZ motor has a factory in Pune, and another enterprise fact indicates that XYZ motor is a supplier to ABC motor, and yet another enterprise fact indicates

that ABC motor is supplier to the enterprise. Further, event information indicates ‘industrial unrest in Pune’. The risk evaluation module correlates these enterprise facts with the event information to determine that the industrial unrest in Pune, where factory of the XYZ motors is located, and the XYZ motor is an entity that indicates an impact on the ABC motor that is directly associated with the supply chain operation of the enterprise. The risk evaluation module therefore analyzes the impact of the event on the one or more entities associated with the enterprise, by correlating the enterprise facts with the event information. In an implementation, the risk evaluation module 108 may also refers to a plurality of pre-defined business rules stored within a rule repository, such as rules 224 associated with the enterprise information fusion system 102, to analyze the impact of the event on the one or more entities associated with the enterprise. The risk evaluation module 108 may store the result of the impact analysis in an evaluation repository, such as evaluation data 222 associated with the enterprise information fusion system 102.
[00058] At block 308, risk associated with the event on the one or more operations of
the enterprise is evaluated based on the impact. In operation, if the result of the impact analysis indicates an impact on the one or more entities associated with the enterprise, the risk evaluation module 108 extracts additional enterprise facts associated with the one or more entities under impact. The additional enterprise facts are indicative of the operational and/or transactional details associated with the entities under impact. The risk evaluation module 108 retrieves additional enterprise facts from the enterprise data, such as enterprise data 220. The risk evaluation module 108 may store the additional enterprise facts in the enterprise facts 218. The risk evaluation module 108 correlates the impact analysis result with the additional enterprise facts to evaluate the risk associated with the impact. For example, if the result of the impact analysis indicates an impact on a key-supplier of the enterprise and the additional enterprise facts indicates that there are supplies pending from the impacted key-supplier, such an impact indicates risk to the supply chain operation of the enterprise. In an implementation, the risk evaluation module 108 may also refer to a plurality of pre-defined business rules stored in the rule repository, such as rules 224 for evaluating the risk associated with the impact on the one or more operations of the enterprise.
[00059] At block 310, an alert is generated, based on the evaluation. If the result of the
risk evaluation indicates a risk associated with the impact or the event in general, on the one

or more operations of the enterprise. The risk evaluation module 108 generates an alert indicating the risk associated with the event on the one or more operations of the enterprise, thereby making the enterprise aware of such risks. The enterprise may therefore take preventive or corrective measures to deal with such risks. In an implementation, the generated alerts may also indicate the probability of the risk associated with the event. The risk evaluation module 108 may store the generated alert in an alerts repository, such as alerts 226 associated with the enterprise information fusion system 102. The risk evaluation module 108 may communicate the generated alerts to the enterprise in form of email communications, Short Messaging Service (SMS), posting on the dashboard, etc.
[00060] Although embodiments for enterprise information fusion have been described
in language specific to structural features and/or methods, it is to be understood that the invention is not necessarily limited to the specific features or methods described. Rather, the specific features and methods are disclosed as exemplary implementations for the enterprise information fusion systems and methods.

I/We claim:
1. A computerized method comprising:
obtaining event information corresponding to at least one event, wherein the at least one event includes at least one of an incident and a customer feedback;
retrieving information corresponding to at least one entity associated with at least one operation of an enterprise from a plurality of entity information sources, based on the event information;
determining an impact of the at least one event on the at least one entity;
evaluating a risk associated with the at least one event on the at least one operation of the enterprise, based on the determining; and
generating an alert based on the evaluating, wherein the alert is indicative of the risk.
2. The method as claimed in claim 1, wherein the at least one operation is a supply chain operation.
3. The method as claimed in claim 1, wherein the at least one entity comprises at least one of a supplier, a vendor, a distributor, and a customer.
4. The method as claimed in claim 1, wherein the determining comprises correlating the event information with the information corresponding to the at least one entity.
5. The method as claimed in claim 1, wherein the evaluating comprises correlating the impact with enterprise data associated with the at least one entity under the impact, wherein the enterprise data is indicative of data related to the at least one operation.
6. A system (102) comprising:
a processor (202); and
a memory (204) coupled to the processor (202), the memory (204) comprising:
an event extraction module (212) configured to,
extract event information corresponding to at least one event,
wherein the at least one event includes at least one of an incident and a
customer feedback; and
a risk evaluation module (108) configured to,

retrieve information corresponding to at least one entity associated with at least one operation of an enterprise from a plurality of entity information sources, based on the event information;
determine an impact of the at least one event on the at least one entity; and
evaluate a risk associated with the at least one event on the at least one operation of the enterprise, based on the determination.
7. The system (102) as claimed in claim 6, wherein the risk evaluation module (108) is further configured to generate an alert based on the evaluation, wherein the alert is indicative of the risk.
8. The system (102) as claimed in claim 6, wherein the at least one operation is a supply chain operation.
9. The system (102) as claimed in claim 6, wherein the at least one entity comprises at least one of a supplier, a vendor, a distributor, and a customer.
10. The system (102) as claimed in claim 6, wherein the risk evaluation module (108) determine the impact by correlating the event information with the information corresponding to the at least one entity.
11. The system (102) as claimed in claim 6, wherein the risk evaluation module (108) evaluate the risk by correlating the impact with enterprise data associated with the at least one entity under the impact, wherein the enterprise data is indicative of data related to the at least one operation.
12. A computer-readable medium having embodied thereon a computer program for executing a method comprising:
obtaining event information corresponding to one or more events, from social media, wherein the one or more events comprises incidents;
retrieving information corresponding to at least one entity associated with the at least one operation of an enterprise from a plurality of entity information sources, based on the event information corresponding to each of the one or more events;
determining an impact of each of the one or more events on the at least one entity, based on the retrieving;

evaluating a risk associated with each of the one or more events on the at least one operation of the enterprise, based on the determining; and
generating an alert based on the evaluation, wherein the alert is indicative of the risk.