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:
“METHOD AND SYSTEM FOR FORECASTING PRODUCT RELATED EVENTS IN A REGION”
2. APPLICANT (S) -
(a) Name : Zensar Technologies Limited
(b) Nationality : Indian
(c)Address : Plot#4 Zensar Knowledge Park, MIDC, Kharadi,
Off Nagar Road, Pune, Maharashtra - 411014, India.
The following specification particularly describes the invention and the manner in which it is to be performed.

TECHNICAL FIELD
[001] The present invention relates to the field of data analysis, and more particularly to analyzing data obtained from various sources in order to forecast particular events in a region.
BACKGROUND OF INVENTION
[002] Forecasting product related events such as demand is an essential tool to successfully cater to the needs of consumers in a number of domains. Under any prevailing market condition, it provides a method to predict the amount of goods and services that might be consumed. Forecast of demand may also affect other aspects of supply chain, right from raw material procurement, manpower estimation, manufacturing, and inventory management by retailers. Hence, it has a cascading effect on every product at all stages of supply chain. It may also be a differentiator in the success and performance of companies in terms of profitability, product availability and customer satisfaction.
[003] A number of factors play a role in forecasting demand of products. Some factors that are observable but not easily quantifiable are events such as social unrest, pandemic, economic recession and the like which may vary from one region to another region. Hence, it goes without saying that multiple other modalities may be an important input to modeling and forecasting exercise, even though they are difficult to quantify. Although, holidays are to a large extent seasonal and readily available as input for modeling, other events cannot be easily identified and factored in the forecast model.
[004] Further, demand for specific products categories change due to various factors and for different time duration. Factors can be as simple as the day of the week or climatic season or may be as complex as geopolitical events or global factors.
[005] For long, demand forecasting has been done by providing historic sales data pertaining to a product as an input to a model which is then analysed by a team of experts to arrive at a forecast for demand in the future. However, since the demand forecast through this technique requires manual intervention, there are high chances that the demand forecast might not be highly accurate. Further, since such techniques rely only on the historic sales

data without considering other important factors such as weather of the region, various economic and political factors, online trends etc. in combination with the historic sales data, the accuracy of such forecast models is questionable.
[006] There is, therefore, a need for a system and method that analyses multi-modal data, i.e., data obtained from various sources and related to various modalities, without any human intervention, to forecast various product related events such as demand for products belonging to different categories.
[007] The information disclosed in this background of the disclosure section is only for enhancement of understanding of the general background of the invention and should not be taken as an acknowledgement or any form of suggestion that this information forms the prior art already known to a person skilled in the art.
SUMMARY OF THE INVENTION
[008] The present disclosure overcomes one or more shortcomings of the prior art and provides additional advantages discussed throughout the present disclosure. Additional features and advantages are realized through the techniques of the present disclosure. Other embodiments and aspects of the disclosure are described in detail herein and are considered a part of the claimed disclosure.
[009] In one embodiment of the present disclosure, a method for forecasting product related events in a region is disclosed. The method comprises capturing weather forecast data of the region. The method further comprises obtaining historic data pertaining to the product related events from a historic data source. The method further comprises categorizing the historic data based on a plurality of time periods. The method further comprises generating a set of keywords for a product for which the product related events is to be forecasted, where the set of keywords are generated based on product characteristics of the product. The method further comprises extracting news data impacting the product related events from one or more sources using the set of keywords. The method further comprises extracting trend data and website data impacting the product related events from the one or more media sources, where the news data, the trend data and the website data forms a

plurality of media content. The method further comprises improving a signal-to noise ratio of the plurality of media content. To improve the signal-to-noise ratio, the method comprises generating a plurality of correlation indexes for each of the plurality of media content, where each correlation index indicates a degree of correlation between the product vis-a-vis the media content. The method for improving the signal-to-noise ratio further comprises generating a plurality of media relevance scores corresponding to the plurality of media content, where each media relevance score indicates a level of impact of corresponding media content on the product related events. And, selecting one or more relevant media content, amongst the plurality of media content, based on the plurality of correlation indexes and the plurality of media relevance scores. Further, the method comprises analyzing the weather forecast data, the one or more relevant media content and the historic data after being categorized in order to forecast the product related events for a pre-defined time period.
[0010] In one embodiment of the present disclosure, a system for forecasting product related events in a region is disclosed. The system comprises a receiving unit configured to capture weather forecast data of the region and obtain historic data pertaining to the product related events from a historic data source. The system further comprises a categorization unit configured to categorize the historic data based on a plurality of time periods. The system further comprises a keyword generation unit configured to generate a set of keywords for a product for which the product related events is to be forecasted, where the set of keywords are generated based on product characteristics of the product. The system further comprises an extraction unit configured to extract news data impacting the product related events from one or more sources using the set of keywords and extract trend data and website data impacting the product related events from the one or more media sources, where the news data, the trend data and the website data forms a plurality of media content. The system further comprises a refinement unit configured to improve a signal-to noise ratio of the plurality of media content. The refinement unit comprises a correlation index generation unit configured to generate a plurality of correlation indexes for each of the plurality of media content, where each correlation index indicates a degree of correlation between the product vis-a-vis the media content. The refinement unit further comprises a score

generation unit configured to generate a plurality of media relevance scores corresponding to the plurality of media content, where each media relevance score indicates a level of impact of corresponding media content on the product related events. The refinement unit further comprises a selection unit configured to select one or more relevant media content, amongst the plurality of media content, based on the plurality of correlation indexes and the plurality of media relevance scores. Further, the system comprises an analysis unit configured to analyse the weather forecast data, the one or more relevant media content and the historic data after being categorized in order to forecast the product related events for a pre-defined time period.
[0011] The foregoing summary is illustrative only and is not intended to be in any way limiting. In addition to the illustrative aspects, embodiments, and features described above, further aspects, embodiments, and features will become apparent by reference to the drawings and the following detailed description.
BRIEF DESCITPION OF DRAWINGS
[0012] The embodiments of the disclosure itself, as well as a preferred mode of use, further objectives and advantages thereof, will best be understood by reference to the following detailed description of an illustrative embodiment when read in conjunction with the accompanying drawings. One or more embodiments are now described, by way of example only, with reference to the accompanying drawings in which:
[0013] Figure 1 shows an exemplary environment 100 of a system for forecasting product related events in a region, in accordance with an embodiment of the present disclosure;
[0014] Figure 2 shows a block diagram 200 illustrating a system for forecasting product related events in a region, in accordance with an embodiment of the present disclosure;
[0015] Figure 3 shows a method 300 for forecasting product related events in a region, in accordance with an embodiment of the present disclosure; and
[0016] Figure 4 shows a block diagram of an exemplary computer system 400 for implementing the embodiments consistent with the present disclosure.

[0017] The figures depict embodiments of the disclosure for purposes of illustration only. One skilled in the art will readily recognize from the following description that alternative embodiments of the structures and methods illustrated herein may be employed without departing from the principles of the disclosure described herein.
DETAILED DESCRIPTION OF THE INVENTION
[0018] The foregoing has broadly outlined the features and technical advantages of the present disclosure in order that the detailed description of the disclosure that follows may be better understood. It should be appreciated by those skilled in the art that the conception and specific embodiment disclosed may be readily utilized as a basis for modifying or designing other structures for carrying out the same purposes of the present disclosure.
[0019] The novel features which are believed to be characteristic of the disclosure, both as to its organization and method of operation, together with further objects and advantages will be better understood from the following description when considered in connection with the accompanying figures. It is to be expressly understood, however, that each of the figures is provided for the purpose of illustration and description only and is not intended as a definition of the limits of the present disclosure.
[0020] Disclosed herein is a system and method for forecasting product related events in a region. In the highly competitive world of today, it is very important for organizations to understand the market demand and accurately forecast the amount of goods and services that might be consumed. The demand forecast not just affects the profitability of the organization but also affects various aspects of the supply chain, right from raw material procurement, manpower estimation, manufacturing, and inventory management by the retailers associated with the organization. Hence, it has a cascading effect on every product at all stages of supply chain. Taking for example, an organization such as PhilipsTM that is into the manufacturing of various electronic products belonging to different categories such as lighting, personal care, household products, sound and vision, health, automotive etc. Under the household products category, PhilipsTM manufactures air purifiers and humidifiers. Now, it would be of high importance for PhilipsTM to forecast a demand for

Air Purifiers in a specific region, say for example, Delhi NCR, by not just examining the historic sales data for Delhi NCR but by also analysing data obtained from various other sources such as weather forecast data, online trends and data obtained from various news articles and online retail websites. The combination of data obtained from all the sources may thereby help them in accurately forecasting the demand for the air purifiers.
[0021] For instance, it is generally seen that in Delhi NCR region, the demand for air purifiers increases rapidly during early onset of winter due to stubble burning in neighboring states such as Punjab and Haryana. The event of stubble burning has been a recurring event for many years. The demand may, therefore, be forecasted by analysing various news articles that talk about the poor air quality in Delhi NCR, online trends that show an increase in the purchasing of air purifiers and various online retail websites that provide various offers on the air purifiers. Therefore, through a combinatorial analysis of the data obtained from such sources an accurate forecast may be obtained. Further, it may also help the organization to understand the competition and make improvements to their product accordingly, in order to beat the competition. Further, for certain non-recurring or non-predictive events such as pandemics, natural calamities etc., historic sales data alone cannot suffice in providing accurate demand forecast. For instance, if there is a forest fire in a region, the air quality of the surrounding regions would become highly poor for many days and even months depending upon the severity of the forest fire. Now, such a situation cannot be predicted in advance and therefore, relying only on the historic sales data may in no way provide an accurate forecast for the demand of air purifiers in the regions affected by the forest fire. There is, therefore, a need to rely upon other sources such as news articles, online trends and purchasing patterns on e-retail websites that would allow the organization to forecast the demand for air purifiers in the upcoming days.
[0022] The system disclosed herein therefore, analyses multi-modal data by employing artificial intelligence to forecast a product demand in the near future. The system also provides highly accurate and time-efficient predictions by removing any errors that may arise due to human intervention during the task of analysing such high-volumes of data obtained

from various sources. The detailed working of the system has been explained in the upcoming paragraphs.
[0023] Figure 1 shows an exemplary environment 100 of a system for forecasting product related events in a region, in accordance with an embodiment of the present disclosure. It must be understood to a person skilled in art that the system may also be implemented in various environments, other than as shown in Fig. 1.
[0024] The detailed explanation of the exemplary environment 100 is explained in conjunction with Figure 2 that shows a block diagram 200 of a system 102 for forecasting product related events in a region, in accordance with an embodiment of the present disclosure. Although the present disclosure is explained considering that the system 102 is implemented on a server, it may be understood that the system 102 may be implemented in a variety of computing systems, such as a laptop computer, a desktop computer, a notebook, a workstation, a mainframe computer, a server, a network server, a cloud-based computing environment.
[0025] In one implementation, the system 102 may comprise an I/O interface 202, a processor 204, a memory 206 and the units 208. The memory 206 may be communicatively coupled to the processor 204 and the units 208. Further, the memory 208 may store multi-modal data 206A and a pre-trained model 206B. The significance and use of each of the stored quantities is explained in the upcoming paragraphs of the specification. The processor 204may 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 204 is configured to fetch and execute computer-readable instructions stored in the memory 206. The I/O interface 202 may include a variety of software and hardware interfaces, for example, a web interface, a graphical user interface, and the like. The I/O interface 202 may enable the system 102 to communicate with other computing devices, such as web servers and external data servers (not shown). The I/O interface 202 may facilitate multiple communications within a wide variety of networks and protocol types, including wired networks, for example, LAN, cable, etc., and wireless

networks, such as WLAN, cellular, or satellite. The I/O interface 202 may include one or more ports for connecting many devices to one another or to another server.
[0026] In one implementation, the units 208 may comprise a receiving unit 210, a categorization unit 212, a keyword generation unit 214, an extraction unit 216, a refinement unit 218 and an analysis unit 220. The refinement unit 218 further comprises a correlation index generation unit 218A, a score generation unit 218B and a selection unit 218C. According to embodiments of present disclosure, these units 210-220 may comprise hardware components like processor, microprocessor, microcontrollers, application-specific integrated circuit for performing various operations of the system 102. It must be understood to a person skilled in art that the processor 204 may perform all the functions of the units 210-220 according to various embodiments of the present disclosure.
[0027] Now referring to figure 1, the environment 100 shows a system 102 that receives multi-modal data 104-112 from various sources and analyses the received multi-modal data 104-112 to provide a forecast 114 for a pre-defined time period. In one embodiment, the forecast 114 may be provided in the form of an index. In one embodiment, the pre-defined time period may include at least one of day, week, month and year or any predefine time period. In one embodiment, the daily, weekly, monthly and yearly forecast may be represented in the form of a chart, trends or a simple table. The working of the system 102 is explained with the help of an example of an organization ABC Coolers that wants to forecast the demand for air-conditioners in a specific region such as Pune (India) for the month of May 2021. Now, the city of Pune is known to have a moderate climate where the use of air-conditioners in homes was not common. But, for the past few years, the climate of Pune has changed drastically owing to many factors such as urbanization, global warming, industrialization etc., with the summer months witnessing a high around 38o C or even more. This change in weather has also seen a rapid increase in the demand for air-conditioners. To make this forecast, the receiving unit 210 receives weather forecast data 104 for Pune city during the month of May 202` from various weather information sources including but not limited to National Oceanic and Atmospheric Administration (NOAA), World Weather Information Service (WWIS), Indian Meteorological Department (IMD) etc. The weather forecast data 104 comprises information related to predicted maximum

and minimum temperatures during each day for the month of May 2021, the humidity level and precipitation probability. Further, apart from receiving weather forecast data 104, the weather information sources may also be used to provide historic weather data. For instance, in accordance with the exemplary environment, since ABC Coolers wants to forecast the demand for air-conditioners in May 2021, it may also extract the weather data for the month of May for the previous years in order to make a comparison and arrive at a suitable conclusion.
[0028] The receiving unit 210 further receives historic data 106 pertaining to product related events from a historic data source. In one embodiment, the historic data 106 pertaining to product related events correspond to historic sales data for air-conditioners sold by ABC Purifier during May 2020 in Pune. The categorization unit 212 categorizes the received historic data 106 based on a plurality of time periods. In one embodiment, the categorization unit 212 performs either a day-wise, a week-wise, a month-wise or a year-wise encoding of the historic data 106. In accordance with the exemplary environment, the historic data 106 obtained for the month of May 2020 is encoded in a day-wise format by the categorization unit 212 to obtain a cyclic pattern of the historic data 106. In one embodiment, the encoding to obtain the cyclic pattern of the historic data 106 is performed by using periodic encoding.
[0029] Now, the content from one or more media sources such as news articles and online blogs 108, e-retail websites 110 and online trends 112 enables the system 102 to accurately forecast the demand for a product. However, before the system 102 analyzes the data, it is very important, that the system 102 extracts the data obtained from these sources. For this purpose, the system 102 comprises a keyword generation unit 214 that generates a set of keywords in order to extract news data 108 from the news articles and online blogs 108. In accordance with the exemplary environment, the set of keywords may include “air-conditioner”, “temperature”, “hot”, “rise”, “sale”, “offer” “Pune” etc. Based on the generated keywords, the extraction unit 216 extracts news data 108 from the news articles and online blogs 108 by employing a Natural Language Processing (NLP) technique. The extraction unit 216 further extracts online trend data 112 and website data 110 from the one or more media sources. In one embodiment, the website data 110 may comprise data

obtained from AmazonTM, FlipkartTM etc. Further, in one embodiment, the online trend data 112 may be obtained from GoogleTM trends. In accordance with the exemplary embodiment, let us say that 100 news articles are extracted by the extraction unit 216 basedon the generated set of keywords. In one embodiment, the weather forecast data 104, the historic data 106 the news data 108, the website data 110 and the online trends data 112 is stored in the memory 206 as multi-modal data 206A.
[0030] Now, based on the generated set of keywords a large amount of data may be obtained from the news articles and blogs 108.Further, the data obtained from the online trends 112 and the e-retail websites 112 is also extensive. Therefore, it is customary to refine the pluralityof media content obtained from the sources 108, 110, 112 before analysing in order to improve a signal-to-noise ratio. In order words, the plurality of media content extracted is filtered so as to obtain one or more relevant media content that are highly important in determining the impact of the media content on the demand of the product. This way, a lot of unwanted data is filtered out which further helps the system 102 in performing the analysis at faster rate. In other words, the system’s 102 computation speed is improved while performing the forecasting as only the relevant media content is taken into consideration.
[0031] The improvement in the signal-to-noise ratio is performed by the refinement unit 218 which further comprises the correlation index generation unit 218A, the score generation unit 218B and the selection unit 218C. The correlation index generation unit 218 generates a plurality of correlation indexes for each of the plurality of media content, where each correlation index indicates a degree of correlation between the product vis-a-vis the media content. Now, in accordance with the exemplary environment, since 100 news articles are generated for the air-conditioner, the correlation index generation unit 218A generates a correlation index for each of the 100 news articles that indicates how closely the article is related to the air-conditioner. Further, in a similar fashion a plurality of correlation indexes is generated for the plurality of media content obtained from e-retail websites 110 and online trends 112. In one embodiment, the correlation indexes may be generated in the form of numeric values.

[0032] Further, the score generation unit 218B generates a plurality of media relevance scores for each of the plurality of media content, where each media relevance score indicates a level of impact of media content on the demand of the product. Now, in accordance with the exemplary environment, since 100 news articles are generated for the air-conditioner, the score generation unit 218B generates a media relevance score for each of the 100 news articles that indicates how much impact will the article have on the demand of the air-conditioner. Further, in a similar fashion a plurality of media relevance scores is generated for the plurality of media content obtained from e-retail websites 110 and online trends 112. In one embodiment, the media relevance scores may be generated in the form of numeric values. Once, the plurality of correlation indexes and the plurality of relevance scores are generated, the selection unit 218C selects one or more relevant media content such that each of the one or more relevant media content has a media relevance score greater than a threshold media relevance score and a correlation index greater than a threshold correlation index. In one embodiment, the threshold correlation index and the threshold media relevance score may be pre-defined in the system 102. In accordance with the exemplary embodiment, out of the 100 news articles, 20 news articles are selected for which the correlation index is greater than the threshold correlation index and the media relevance score is greater than the threshold media relevance score. The selected 20 news articles are embedded in the form of a vector. In one embodiment, the vector may be a 64-dimensional vector.
[0033] Further, the selected one or more relevant media content is analysed by the analysis unit 220 by applying a pre-trained model 206B that is stored in the memory 206. In one embodiment, the pre-trained model 206B is at least one of a mid-fusion model and a late-fusion model. In the mid-fusion model, the parameters obtained from individual AI networks concatenating the parameters of individual AI networks which is followed by subsequent multi-layer parameter learning. In contrast, the late-fusion model concatenates parameters in the final layer before making the forecast. The selection of the model depends on the relevance of the modality. For instance, date encoding and the online trends 112 are highly relevant as they continuously associate and add meaningful information to the historical data 106. Hence, mid-fusion model is more suitable. In contrast, late fusion is

used with news articles 108 as they directly associate a relevance factor to the product related events and hence can be concatenated just before making the forecast 114.
[0034] Based on the analysis, in accordance with the exemplary embodiment, a demand forecast 114 for the ABC Coolers’ air-conditioner for the month of May 2021 in Pune city is provided by the I/O interface 202. Further, in one embodiment, the demand forecast 114 may be provided for each day in May 2021 or for each-week of May 2021.
[0035] It may be understood by a skilled person, that the exemplary embodiment depicted in figure 1 focusses on a single product belonging to a particular category. However, the demand forecast may be provided for different products or different categories such as automotive, beauty, bakery, dairy, personal care etc.
[0036] Figure 3 depicts a method 300 for forecasting product related events in a region, in accordance with an embodiment of the present disclosure.
[0037] As illustrated in figure 3, the method 300 includes one or more blocks illustrating a method for forecasting product related events in a region. The method 300 may be described in the general context of computer executable instructions. Generally, computer executable instructions may include routines, programs, objects, components, data structures, procedures, modules, and functions, which perform specific functions or implement specific abstract data types.
[0038] The order in which the method 300 is described is not intended to be construed as a limitation, and any number of the described method blocks may be combined in any order to implement the method. Additionally, individual blocks may be deleted from the methods without departing from the spirit and scope of the subject matter described.
[0039] At block 302, the method 300 may include capturing weather forecast data of the region.
[0040] At block 304, the method 300 may include obtaining historic data pertaining to the product related events from a historical data source.

[0041] At block 306, the method 300 may include categorizing the historic data based on a plurality of time periods.
[0042] At block 308, the method 300 may include generating a set of keywords for a product for which the product related event is to be forecasted. The set of keywords are generated based on product characteristics of the product.
[0043] At block 310, the method 300 may include extracting news data impacting the product related events from one or more media sources using the set of keywords.
[0044] At block 310A, the method 300 may include extracting trend data 110 and website data 112 impacting the product related events from the one or more media sources. The news data 108, the online trend data 112 and the website data 110 forms a plurality of media content.
[0045] At block 312, the method 300 may include improving a signal-to noise ratio of the plurality of media content. To improve the signal-to-noise ratio, the method follows steps 312A-312C.
[0046] At block 312A, the method 300 may include generating a plurality of correlation indexes for each of the plurality of media content. Each correlation index indicates a degree of correlation between of the product vis-a-vis the media content.
[0047] At block 312B, the method 300 may include generating a plurality of media relevance scores corresponding to the plurality of media content. Each media relevance score indicates a level of impact of corresponding media content on the product related events.
[0048] At block 312C, the method 300 may include selecting one or more relevant media content, amongst the plurality of media content, based on the plurality of correlation indexes and the plurality of media relevance scores.

[0049] At block 314, the method 300 may include analyzing the weather forecast data, the one or more relevant media content and the historic data after being categorized in order to forecast the product related events for a pre-defined time period.
Computer System
[0050] Figure 4 illustrates a block diagram of an exemplary computer system 400 for implementing embodiments consistent with the present disclosure. It may be understood to a person skilled in art that the computer system 400 and its components is similar to the system 102 referred in Fig. 2. In an embodiment, the computer system 400 may be a peripheral device, which is used for forecasting product related events in a region. The computer system 400 may include a central processing unit (“CPU” or “processor”) 402. The processor 402 may comprise at least one data processor for executing program components for executing user or system-generated processes. The processor 402 may include specialized processing units such as integrated system (bus) controllers, memory management control units, floating point units, graphics processing units, digital signal processing units, etc.
[0051] The processor 402 may be disposed in communication with one or more input/output (I/O) devices via I/O interface 401. The I/O interface 401 may employ communication protocols/methods such as, without limitation, audio, analog, digital, stereo, IEEE-1394, serial bus, Universal Serial Bus (USB), infrared, PS/2, BNC, coaxial, component, composite, Digital Visual Interface (DVI), high-definition multimedia interface (HDMI), Radio Frequency (RF) antennas, S-Video, Video Graphics Array (VGA), IEEE 802.n /b/g/n/x, Bluetooth, cellular (e.g., Code-Division Multiple Access (CDMA), High-Speed Packet Access (HSPA+), Global System For Mobile Communications (GSM), Long-Term Evolution (LTE) or the like), etc. Using the I/O interface, the computer system 400 may communicate with one or more I/O devices.
[0052] In some embodiments, the processor 402 may be disposed in communication with a communication network 414 via a network interface 403. The network interface 403 may communicate with the communication network 414. The communication unit may employ connection protocols including, without limitation, direct connect, Ethernet (e.g., twisted

pair 10/100/1000 Base T), Transmission Control Protocol/Internet Protocol (TCP/IP), token ring, IEEE 802.11a/b/g/n/x, etc.
[0053] The communication network 414 may be implemented as one of the several types of networks, such as intranet or Local Area Network (LAN) and such within the organization. The communication network 414 may either be a dedicated network or a shared network, which represents an association of several types of networks that use a variety of protocols, for example, Hypertext Transfer Protocol (HTTP), Transmission Control Protocol/Internet Protocol (TCP/IP), Wireless Application Protocol (WAP), etc., to communicate with each other. Further, the communication network 414 may include a variety of network devices, including routers, bridges, servers, computing devices, storage devices, etc.
[0054] In some embodiments, the processor 402 may be disposed in communication with a memory 405 (e.g., RAM 412, ROM 413, etc. as shown in FIG. 4) via a storage interface 404. The storage interface 404 may connect to the memory 405 including, without limitation, memory drives, removable disc drives, etc., employing connection protocols such as Serial Advanced Technology Attachment (SATA), Integrated Drive Electronics (IDE), IEEE-1394, Universal Serial Bus (USB), fiber channel, Small Computer Systems Interface (SCSI), etc. The memory drives may further include a drum, magnetic disc drive, magneto-optical drive, optical drive, Redundant Array of Independent Discs (RAID), solid-state memory devices, solid-state drives, etc.
[0055] The memory 405 may store a collection of program or database components, including, without limitation, user /application, an operating system, a web browser, mail client, mail server, web server and the like. In some embodiments, computer system may store user /application data, such as the data, variables, records, etc. as described in this invention. Such databases may be implemented as fault-tolerant, relational, scalable, secure databases such as OracleR or SybaseR.
[0056] The operating system may facilitate resource management and operation of the computer system. Examples of operating systems include, without limitation, APPLE MACINTOSHR OS X, UNIXR, UNIX-like system distributions (E.G., BERKELEY

SOFTWARE DISTRIBUTIONTM (BSD), FREEBSDTM, NETBSDTM, OPENBSDTM,
etc.), LINUX DISTRIBUTIONSTM (E.G., RED HATTM, UBUNTUTM,
KUBUNTUTM, etc.), IBMTM OS/2, MICROSOFTTM WINDOWSTM (XPTM,
VISTATM/7/8, 10 etc.), APPLER IOSTM, GOOGLER ANDROIDTM,
BLACKBERRYR OS, or the like. A user interface may facilitate display, execution, interaction, manipulation, or operation of program components through textual or graphical facilities. For example, user interfaces may provide computer interaction interface elements on a display system operatively connected to the computer system, such as cursors, icons, check boxes, menus, windows, widgets, etc. Graphical User Interfaces (GUIs) may be employed, including, without limitation, APPLE MACINTOSHR operating systems, IBMTM OS/2, MICROSOFTTM WINDOWSTM (XPTM, VISTATM/7/8, 10 etc.), UnixR X-Windows, web interface libraries (e.g., AJAXTM, DHTMLTM, ADOBE® FLASHTM, JAVASCRIPTTM, JAVATM, etc.), or the like.
[0057] Furthermore, one or more computer-readable storage media may be utilized in implementing embodiments consistent with the present invention. A computer-readable storage medium refers to any type of physical memory on which information or data readable by a processor may be stored. Thus, a computer-readable storage medium may store instructions for execution by one or more processors, including instructions for causing the processor(s) to perform steps or stages consistent with the embodiments described herein. The term “computer-readable medium” should be understood to include tangible items and exclude carrier waves and transient signals, i.e., non-transitory. Examples include Random Access Memory (RAM), Read-Only Memory (ROM), volatile memory, non-volatile memory, hard drives, Compact Disc (CD) ROMs, Digital Video Disc (DVDs), flash drives, disks, and any other known physical storage media.
[0058] A description of an embodiment with several components in communication with each other does not imply that all such components are required. On the contrary, a variety of optional components are described to illustrate the wide variety of possible embodiments of the invention.
[0059] When a single device or article is described herein, it will be clear that more than one device/article (whether they cooperate) may be used in place of a single device/article.

Similarly, where more than one device or article is described herein (whether they cooperate), it will be clear that a single device/article may be used in place of the more than one device or article or a different number of devices/articles may be used instead of the shown number of devices or programs. The functionality and/or the features of a device may be alternatively embodied by one or more other devices which are not explicitly described as having such functionality/features. Thus, other embodiments of the invention need not include the device itself.
[0060] Finally, the language used in the specification has been principally selected for readability and instructional purposes, and it may not have been selected to delineate or circumscribe the inventive subject matter. It is therefore intended that the scope of the invention be limited not by this detailed description, but rather by any claims that issue on an application based here on. Accordingly, the embodiments of the present invention are intended to be illustrative, but not limiting, of the scope of the invention, which is set forth in the following claims.
[0061] While various aspects and embodiments have been disclosed herein, other aspects and embodiments will be apparent to those skilled in the art. The various aspects and embodiments disclosed herein are for purposes of illustration and are not intended to be limiting, with the true scope and spirit being indicated by the following claims.

Reference Numerals:
Reference Numeral Description
Exemplary environment of a system for forecasting product related
100
events in a region
102 System
104 Weather Forecast Data
106 Historic Data
108 News Articles
110 Website Data
112 Online Trends
114 Forecast
200 Block diagram of the system
202 I/O Interface
204 Processor
206 Memory
206A Multi-modal data
206B Pre-trained model
208 Units
212 Receiving Unit
214 Categorization Unit
216 Extraction Unit
218 Refinement Unit
218A Correlation Index Unit

218B Score Generation Unit
218C Selection Unit
220 Analysis Unit

We Claim:
1. A method for forecasting product related events in a region, the method comprises:
obtaining (304) historic data (106) pertaining to the product related events from a historical data source;
categorizing (306) the historic data (106) based on a plurality of time periods;
generating (308) a set of keywords for a product for which the product related events is to be forecasted, wherein the set of keywords are generated based on product characteristics of the product;
extracting (310) news data (108) impacting the product related events from one or more media sources using the set of keywords;
extracting (310A) online trend data (110) and website data (112) impacting the product related events from the one or more media sources, wherein the news data (108), the online trend data (110) and the website data (112) forms a plurality of media content;
improving (312) a signal-to-noise ratio of the plurality of media content (108, 110, 112) by:
generating (312A) a plurality of correlation indexes for each of the plurality
of media content, wherein each correlation index indicates a degree of correlation
between the product vis-a-vis the media content;
generating (312B) a plurality of media relevance scores corresponding to
the plurality of media content, wherein each media relevance score indicates a level
of impact of corresponding media content on the product related events;
selecting (312C) one or more relevant media content, amongst the plurality
of media content, based on the plurality of correlation indexes and the plurality of
media relevance scores;
analyzing (314) the one or more relevant media content and the historic data after being categorized in order to forecast the product related events for a pre-defined time period.
2. The method (300) as claimed in claim 1, further comprising:
capturing (302) weather forecast data in the region; and
applying a pre-trained model (206B) upon the weather forecast data (104), the one or more relevant media content, and the historic data after being categorized to forecast the product related events, wherein the pre-trained model (206B) is at least one of a mid-fusion model and a late-fusion model.
3. The method (300) as claimed in claim 1, wherein the one or more media sources comprises
at least one of news articles and online blogs (108), online trends (110) and websites (112).

4. The method (300) as claimed in claim 1, wherein a relevant media content of the one or more relevant media content obtained from the news articles is represented in the form of a vector embedding.
5. The method (300) as claimed in claim 1, wherein categorizing (306) the historic data (106) based on the plurality of time periods further comprises performing at least one of:
a day-wise, a week-wise, a month-wise and a year-wise encoding of the historic data (106).
6. The method (300) as claimed in claim 1, wherein each of the one or more relevant media content selected has a media relevance score greater than a threshold media relevance score and a correlation index greater than a threshold correlation index.
7. A system (102) for forecasting product related events in a region, the system comprises:
a receiving unit (210) configured to obtain historic data (106) pertaining to the product related events from a historical data source;
a categorization unit (212) configured to categorize the historic data (106) based on a plurality of time periods;
a keyword generation unit (214) configured to generate a set of keywords for a product for which the product related events is to be forecasted, wherein the set of keywords are generated based on product characteristics of the product;
an extraction unit (216) configured to:
extract news data (108) impacting the product related events from one or
more media sources using the set of keywords;
extract trend data (112) and website data (110) impacting the product related
events from the one or more media sources, wherein the news data (108), the online
trend data (112) and the website data (110) forms a plurality of media content;
a refinement unit (218) configured to improve a signal-to noise ratio of the plurality of media content, wherein the refinement unit further comprises:
a correlation index generation unit (218A) configured to generate a plurality
of correlation indexes for each of the plurality of media content, wherein each
correlation index indicates a degree of correlation between the product vis-a-vis the
media content;
a score generation unit (218B) configured to generate a plurality of media
relevance scores corresponding to the plurality of media content, wherein each
media relevance score indicates a level of impact of corresponding media content
on the product related events;
a selection unit (218C) configured to select one or more relevant media
content, amongst the plurality of media content, based on the plurality of correlation
indexes and the plurality of media relevance scores; and

an analysis unit (220) configured to analyze the one or more relevant media content and the historic data after being categorized in order to forecast the product related events for a pre-defined time period.
8. The system (102) as claimed in claim 7, wherein:
the receiving unit (210) is further configured to capture weather forecast data (104) of the region; and
the analysis unit (220) is further configured to apply a pre-trained model (206B) upon the weather forecast data, the one or more relevant media content, and the historic data after being categorized in order to forecast the product related events, wherein the pre-trained model (206B) is at least one of a mid-fusion model and a late-fusion model.
9. The system (102) as claimed in claim 7, wherein the one or more media sources comprises at least one of news articles (108), online trends (112) and websites (110), and wherein the weather forecast data (104) is captured from one or more weather information sources.
10. The system (102) as claimed in claim 7, wherein a relevant media content of the one or more relevant media content obtained from the news articles is represented in the form of a vector embedding.
11. The system (102) as claimed in claim 7, wherein to categorize the historic data (106) based on the plurality of time periods, the categorization unit (212) is further configured to perform at least one of:
a day-wise, a week-wise, a month-wise and a year-wise encoding of the historic data.
12. The system (102) as claimed in claim 7, wherein the selection unit (218C) selects the one
or more relevant media content such that each of the one or more relevant media content
has a media relevance score greater than a threshold media relevance score and a
correlation index greater than a threshold correlation index.