Patent Application
20170372135
In many contexts, such as the service industry, work is generally organized into processes that often entail printing documents. There is a growing trend towards replacing printing paper documents with digital counterparts, which may entail use of electronic signatures, email (instead of post mail) and online form filling. There are many reasons for this change, including higher productivity, cost-efficiency, and becoming more environmentally-friendly. Many large organizations are therefore looking for solutions to reduce paper usage and to move from using paper to digital documents. Unfortunately, especially in large organizations, it is often difficult to achieve this goal, because of a lack of information. Those in management, for example, often do not have a detailed understanding of where paper is being used by company employees, in particular, in which tasks or subtasks paper documents are generated, as well as how much paper is used in the process, in terms of the volume of paper being used in each of these tasks. Nor is there a good understanding of the reasons why paper is used for these tasks, i.e., what are the barriers that prevent using digital versions instead of paper documents within these tasks.
Having answers to these questions would help organizations to select which processes/tasks could be modified to facilitate moving them from paper to digital. However, without a good understanding of the paper consumption of the various tasks, and the reasons for printing documents, it is difficult to focus these efforts on the processes where changes would be the most effective.
It is now becoming important to not only looking at ways to facilitate printing inside a client corporation, but as well at optimizing printing by replacing inefficient paper workflows by more efficient electronic ones. The reasons for printing documents are often task dependent. Some common reasons involve requiring signatures, archiving, transitions between different computer systems, crossing organizational barriers, and so forth. However, there may be other reasons that have not been identified by the organization. To move from paper to digital, appropriate solutions may need to be implemented to replace the functions previously provided through generating paper documents, such as digital archiving, digital signatures, and the like. However, for some tasks, paper may afford benefits that digital documents do not provide. Paper is, for example, easy portable (e.g., when traveling), easy to read and annotate, and easy to hand over to another person. Employees could be provided with portable devices, such as eReaders, to address some of these issues, but this solution may not be cost-effective.
In this context, consultants are currently able to analyze how and what employees print within a client corporation, to infer associated workflows and to suggest well adapted replacement solutions, reducing paper usage and increasing productivity. Therefore, consultants are currently collecting print volume information directly from the devices and the estimated time spent per employee on the different tasks or processes through a survey. They furthermore conduct individual interviews with selected particularly paper intensive employees to get a deeper understanding of their paper processes.
However, from a human point of view, it is difficult to motivate those employees to free time for talking about their print usage, in other words, about their ways of working. Indeed, this topic is often not motivating and fuzzy. Finally, the survey and interview approach also demands a lot of time from the consultant to identify and suggest processes to optimize. Furthermore, the consultant's proposals are often rather inspired by his prior experience with other companies than guided by the information collected in the target corporation. Thus, the consultant usually first concentrates on well-known ubiquitous standard processes and [semi] structured workflows, and often misses less frequent or less typical unstructured and hidden workflows that nevertheless exist in every work place.
There remains a need for a system and method of identifying unusual paper-intensive workflows in a more efficient, open, accurate and motivating fashion, with a need to gather employee knowledge and to combine it with machine learning techniques in a short term and collaborative workshop.
The following references, the disclosures of which are incorporated herein by reference in their entireties, are mentioned:
U.S. patent application Ser. No. 14/607,739, filed Jan. 28, 2015, by Willamowski et al., and entitled “SYSTEM AND METHOD FOR THE CREATION AND MANAGEMENT OF USER-ANNOTATIONS ASSOCIATED WITH PAPER-BASED PROCESSES”
U.S. Publication No. 2011/0137898, Published Jun. 9, 2011, by Gordo et al., and entitled “UNSTRUCTURED DOCUMENT CLASSIFICATION”;
U.S. Pat. No. 7,366,705, Issued Apr. 29, 2008, by Zeng et al., and entitled “CLUSTERING BASED TEXT CLASSIFICATION”;
U.S. Pat. No. 8,165,410, Issued Apr. 24, 2012, by Perronnin and entitled “BAGS OF VISUAL CONTEXT-DEPENDENT WORDS FOR GENERIC VISUAL CATEGORIZATION”;
U.S. Pat. No. 8,280,828, issued Oct. 2, 2012, by Perronnin et al., and entitled “FAST AND EFFICIENT NONLINEAR CLASSIFIER GENERATED FROM A TRAINED LINEAR CLASSIFIER”;
U.S. Pat. No. 8,532,399, Issued Sep. 10, 2013, by Perronnin et al., and entitled “LARGE SCALE IMAGE CLASSIFICATION”;
U.S. Pat. No. 8,731,317, issued May 20, 2014, by Sanchez et al., and entitled “IMAGE CLASSIFICATION EMPLOYING IMAGE VECTORS COMPRESSED USING VECTOR QUANTIZATION”;
U.S. Pat. No. 8,879,103, by Willamowski et al., Issued Nov. 4, 2014 and entitled “SYSTEM AND METHOD FOR HIGHLIGHTING BARRIERS TO REDUCING PAPER USAGE”; and
CSURKA et al., “WHAT IS THE RIGHT WAY TO REPRESENT DOCUMENT IMAGES?”, Xerox Research Center Europe, Grenoble, France, Mar. 25, 2016, pages 1-35, are incorporated herein by reference in their entirety.
In one embodiment of this disclosure, described is a computer-implemented method for gathering knowledge related to paper-intensive processes associated with one or more printing systems used in an organization to generate printed documents by a group of users. The method comprises generating a representative set of printed documents by tracking and storing all or a portion of the printed documents and associated metadata generated by the printing system over a predetermined duration of time and processing the representative set of printed documents to generate a plurality of clusters of printed documents, each cluster of printed documents including a subset of the representational set of printed documents which are associated with a predefined measurement of similarity. The method then assigns a set of users to label each cluster of printed documents, each set of users including a subset of users selected from the group of users and each subset of users associated with a relatively high degree of contribution to the cluster of printed documents, relative to other users, included in the group of users. After the users have labeled, the method is configured to receive the document labeling data from the subsets of users for one or more printed documents associated with each of the respective document clusters, the labeling data including one or more of a process type, a document type and a reason for printing the printed document. The document labeling information is used to train a classifier using all or part of the received document labeling data and associated printed documents and using the trained classifier, classifying one or more printed documents generated at the beginning of the process but not yet labeled. The method further compiles the label data for all or a portion of the representative set of printed documents, including label data directly provided by one or more users and label data provided by the method and generates one or more indicators representing the use of printed documents associated with one or more of the document type, the process, the user, a project and the reason for printing.
In another embodiment of this disclosure, described is a system for gathering knowledge related to paper-intensive processes associated with one or more printing systems used in an organization to generate printed documents by a group of users. The system comprises a print job tracking component configured to generate a representative set of printed documents by tracking and storing all or a portion of the printed documents and associated metadata generated by the printing system over a predetermined duration of time and a clustering component configured to process the representative set of printed documents to generate a plurality of clusters of printed documents, each cluster of printed documents including a subset of the representational set of printed documents which are associated with a predefined measurement of similarity. The system further includes an annotation component configured to assign a set of users to label each cluster of printed documents, each set of users including a subset of users selected from the group of users and each subset of users associated with a relatively high degree of contribution to the cluster of printed documents, relative to other users, included in the group of users, and receive document labeling data from the subsets of users for one or more printed documents associated with each of the respective document clusters, the labeling data including one or more of a process type, a document type and a reason for printing the printed document. A classifier component of the system is configured to be trained using all or part of the received document labeling data and associated printed documents and classify one or more of the printed documents generated by the print job tracking component which were not included in the plurality of clusters of printed documents generated by the clustering component and a compiler is configured to compile the label data for all or a portion of the representative set of printed documents, including label data directly provided by one or more users and label data provided by the classifier component. Lastly the system includes an indicator generation component configured to generate one or more indicators representing the use of printed documents associated with one or more of the document type, the process, the user, a project and the reason for printing.
In still another embodiment of this disclosure, described is a computer program product comprising a non-transitory recording medium storing instructions which, when executed by a computer processor, perform a method for gathering knowledge related to paper-intensive processes associated with one or more printing systems used in an organization to generate printed documents by a group of users. The method comprises generating a representative set of printed documents by tracking and storing all or a portion of the printed documents and associated metadata generated by the printing system over a predetermined duration of time. The representative set of printed documents is processed to generate a plurality of clusters of printed documents, each cluster of printed documents including a subset of the representational set of printed documents which are associated with a predefined measurement of similarity. A set of users is assigned label each cluster of printed documents where each set of users includes a subset of users selected from the group of users and each subset of users is associated with a relatively high degree of contribution to the cluster of printed documents, relative to other users, included in the group of users. The method receives document labeling data from the subsets of users for one or more printed documents associated with each of the respective document clusters, the labeling data including one or more of a process type, a document type and a reason for printing the printed document. The information is used to train a classifier using all or part of the received document labeling data and associated printed documents and using the trained classifier, classifying one or more representative set of printed documents which were not included in the plurality of clusters of printed documents previously clustered. The label data is compiled for all or a portion of the representative set of printed documents, including label data directly provided by one or more users and then generating one or more indicators representing the use of printed documents associated with one or more of the document type, the process, the user, a project and the reason for printing.
To more effectively gather knowledge about paper-intensive processes in an organization, a system and method for supporting the preparation, animation and execution of a collaborative workshop for high speed and efficient document labeling and workflow assessment is disclosed. The method is structured in several steps and phases, and the system provides different types of support and guidance throughout these various steps and phases. It enables on one hand a human facilitator to efficiently prepare and animate the workshop and to optimally engage the participants. This facilitator role can be fulfilled by an external consultant, specialized in the analysis and improvement of organizational paper-based work processes in general. The system enables on the other hand the individual workshop participants to collaboratively and efficiently label pre-selected documents. These participants can be a small number of selected employees working in the target organization, and the documents to label correspond to a selected set of paper documents produced by the participants in their work. The information provided as part of the workshop from users is then compiled and used to train a classifier to continue the task of labeling documents that remain unlabeled.
According to an exemplary embodiment the disclosed method and system structures the workshop organization by animating and executing in steps and phases. Below is described first the different steps and phases and then how the functionalities provided by the system support these individual steps and phases.
The method and system provided includes the following steps:
1. Workshop Preparation:
2. Workshop Execution:
To provide for the different steps and phases the system includes the following processes/methods:
1. Document Clustering and Analysis:
2. Document Categorization:
3. Monitoring Workshop Progression Indicators:
Based on these indicators the system may generate suggestions—either directly or indirectly through the mediation of the facilitator—that one or more participants switch clusters and/or leave/join groups for more efficient labeling and/or that the workshop transitions to the next phase, the requalification phase.
With reference to
A consultant 118 is the facilitator of the experience. The consultant 118 knows how to progress towards the ultimate goal of paper workflow identification and optimization and with the knowledge of previous experiences in other companies, knows how to guide and motivate the participants through a smooth experience. The consultant 118 is assisted by the system which introduces a clear progression metrics and on the fly indicators and guidelines, and is the human representative mediating between the system and the workshop participants.
Users 106 are employees of the target organization who print documents in the context of their work and who have the knowledge about the purpose of their printing. They contribute their individual view of the work processes from their different angles, with respect to their department or role in the company. They are able to grasp and recognize a document they have printed and to explain why they had to print it. They are able, as well, to discuss these points all together to reach a common understanding.
The system 100 includes a print job tracking component 102 that intercepts print jobs 104 that are sent by different users 106 within the organization to a printing infrastructure 108 (and/or which receives information on the print jobs from the printing infrastructure, such as print logs). The print job tracking component is configured to track and store all or a portion of the printed documents and their metadata that is generated by the printing system over a specified period of time. The number of users and print jobs is not limited and each user may generate one or more print jobs for printing on the printing infrastructure 108.
The clustering component 114 identifies clusters 116 of similar print jobs 104. The clustering is based on the assumption that similar print jobs will belong to similar tasks and that users have work roles corresponding to a specific subset of tasks and thus print essentially the corresponding types of print jobs. Thus, print jobs which have no annotations can be clustered based on the similarity of their print job signatures to those of annotated jobs. Each cluster of printed documents includes a subset of the representative set of printed documents generated by the job tracking component.
An annotation component 112 is configured to assign a subset of users to review and label each cluster of printed documents generated by the clustering component. The set of users 106 includes users who have a relatively high degree of contribution to the representative set of printed documents. The annotation component 112 assigns a subset of these users to each cluster based upon the users' 106 relatively high degree of printing of the documents in the given cluster. The annotation component 112 then receives document labeling data from the subsets of users 106 for each cluster for one or more of the printed documents 104. The labeling data received by the annotation component includes one or more of a process type, a document type, and a reason for printing the printed document. A compiler 110 is configured to compile all the received label data for all or a portion of the representative set of printed documents generated by the job tracking component. The label data used by the compiler can be label data provided directly by one or more of the users and label data provided by the classifier component.
A classifier component 242 is configured to be trained using all or portions of the document labeling data as well as using associated printed documents. The classifier component 242 then classifies one or more of the printed documents generated by the print job tracking component 102 which were not included in the plurality of clusters of printed documents 104 generated by the clustering component 114.
A compiler component 110 is configured to compile the label data for all or a portion of the representative set of printed documents 104. The compiler compiles label data that has been directly provided by one or more users 106 as well as label data provided by the classifier component 242.
The system further includes an indicator generation component 244 that is configured to generate one or more indicators representing the use of printed documents associated with the one or more document type, the process, the user, a project, and the reason for printing.
As illustrated in
An analysis component 206 generates task-related information 208, based on the clustering and annotations, which is output from the system 100. In the exemplary embodiment, the components 102, 110, 112, 116, 206 are in the form of software which is implemented by a computer processor 201 in communication with memory 202.
In the illustrated embodiment, the computing device 200 receives print job information comprising print jobs 104, and/or information extracted therefrom, such as print logs 212, via a network. In one embodiment the print jobs 104 are received by the job tracking component 102 from a plurality of client computing devices 214, 216, 218 linked to the network, that are used by the respective users 106 to generate print jobs. However, it is to be appreciated that print job information for the submitted print jobs 104 may alternatively or additionally be received from the printing infrastructure 108 or from a print job server (not shown), which distributes the print jobs 104 to the various printers in printing infrastructure 108. The print job information 104, 212 is received by the system 100 via one or more input/output (I/O) interfaces 220, 222 and stored in data memory 224 of the system 100 during processing. The computing device 200 also may control the distribution of the received print jobs 104 to respective printers 226, 228 of the printing infrastructure 108, or this function may be performed by another computer on the network.
The feature extractor 110 extracts features from the print job information. The extracted features are used to generate a representation 230 of each print job, which may be stored in memory 224.
The annotation component 112 receives, as input, print job annotations 232 for at least some of the print jobs 104, via the network, e.g., from the client computing devices 214, 216, 218 and stores the annotations, or information extracted from them, in memory 112. The annotations may include task-related information and/or information on constraints provided in the form of a note which limit or prevent the user's ability to use a digital version of the printed document rather than printing a paper copy. Alternatively, the task-related information may include a task category selected from a plurality of task categories, or information from which the task category may be inferred. The constraint-related information may include a constraint category selected from a plurality of constraint categories, or information from which the constraint category may be inferred.
The clustering component 114 may be trained, on the annotated (labeled) print jobs and is then able to cluster a set of labeled and unlabeled print jobs into a plurality of clusters 116. Hardware components 202, 210, 220, 222, 224 may communicate via a data/control bus 234. The processor 210 executes the instructions for performing the method outlined in
The client devices 214, 216, 218 may each communicate with one or more of a display 236, for displaying information to users, and a user input device 238, such as a keyboard or touch or writable screen, a cursor control device, such as mouse or trackball, a speech to text converter, or the like, for inputting text and for communicating user input information and command selections to the respective computer processor and to processor 210 via network.
The computer device 200 may be a PC, such as a server computer, a desktop, laptop, tablet, or palmtop computer, a portable digital assistant (PDA), a cellular telephone, a pager, combination thereof, or other computing device capable of executing instructions for performing the exemplary method.
The memory 202, 224 may represent any type of non-transitory computer readable medium such as random access memory (RAM), read only memory (ROM), magnetic disk or tape, optical disk, flash memory, or holographic memory. In one embodiment, the memory 202, 224 comprises a combination of random access memory and read only memory. In some embodiments, the processor 210 and memory 202 may be combined in a single chip. The network interface 220, 222 allows the computer 200 to communicate with other devices via a computer network, such as a local area network (LAN) or wide area network (WAN), or the internet, and may comprise a modulator/demodulator (MODEM) a router, a cable, and/or Ethernet port. Memory 202, 224 stores instructions for performing the exemplary method as well as the processed data 208.
The digital processor 210 can be variously embodied, such as by a single-core processor, a dual-core processor (or more generally by a multiple-core processor), a digital processor and cooperating math coprocessor, a digital controller, or the like. The exemplary digital processor 210, in addition to controlling the operation of the computer 200, executes instructions stored in memory 204 for performing the method outlined in
The client devices 214, 216, 218 may be configured as for computing device 200, except as noted.
The term “software,” as used herein, is intended to encompass any collection or set of instructions executable by a computer or other digital system so as to configure the computer or other digital system to perform the task that is the intent of the software. The term “software” as used herein is intended to encompass such instructions stored in storage medium such as RAM, a hard disk, optical disk, or so forth, and is also intended to encompass so-called “firmware” that is software stored on a ROM or so forth. Such software may be organized in various ways, and may include software components organized as libraries, Internet-based programs stored on a remote server or so forth, source code, interpretive code, object code, directly executable code, and so forth. It is contemplated that the software may invoke system-level code or calls to other software residing on a server or other location to perform certain functions.
As will be appreciated,
With reference to
At S302, print job information 104, 212 is acquired for a collection of print jobs generated by a set of users 106, such as company employees, and stored in computer memory 224. The method generates a set of representative printed documents by tracking and storing all or a portion of the printed documents and their associated metadata generated by the printing system over a predetermined duration of time.
At S304 the representative set of printed documents is processed to generate a plurality of clusters. The clusters contain a subset of printed documents which are associated with a predefined measurement of similarity.
At S306, users of the system are assigned to label the subset of documents included in various clusters. The set of users are selected from a group of users where each user in a group has shown a relatively high contribution to the cluster of printed documents i.e. the user has created or printed a large portion of the documents contained in the cluster.
At S308, user annotations 230 are received by the system 100 and stored in memory.
At S310, using all or a part of the document labeling information received from the users is used to train a classifier.
At S312, using the trained classifier, the set of representative documents that have not yet been classified by the users is classified by the system.
At S314, the label data generated by the users or by the compiler is compiled including the label data directly provided by the one or more users and the label data provided by the compiler.
At S316, the method generates one or more indicators representing the use of the printed documents associated with one or more document type, the process, the user, a project, and the reason for printing. The method ends at S318.
The method illustrated in
Alternatively or additionally, the method may be implemented in transitory media, such as a transmittable carrier wave in which the control program is embodied as a data signal using transmission media, such as acoustic or light waves, such as those generated during radio wave and infrared data communications, and the like.
The exemplary method may be implemented on one or more general purpose computers, special purpose computer(s), a programmed microprocessor or microcontroller and peripheral integrated circuit elements, an ASIC or other integrated circuit, a digital signal processor, a hardwired electronic or logic circuit such as a discrete element circuit, a programmable logic device such as a PLD, PLA, FPGA, Graphical card CPU (GPU), or PAL, or the like. In general, any device, capable of implementing a finite state machine that is in turn capable of implementing the flowchart shown in
Further details of the system and method will now be described.
Print job tracking systems that provide the basic functionality of the exemplary print job tracking component 102, such as intercepting print jobs issued through a print infrastructure and extracting the corresponding user name, document title, document length, and similar information are readily available.
Various procedures for annotation are contemplated which can be used individually or in combination. For example, the annotation process can be initiated spontaneously by the users or when requested by the system, for example, to use active learning in order to validate or refine the actual clustering. Users may annotate one (or a set of) selected print job(s), thereby associating it to a corresponding one of a set of tasks and identifying constraints on printing. In another embodiment, the user may annotate a point in time or time frame with the task they were mainly performing at that time (e.g., reviewing papers for a conference, preparing for a customer visit, etc.) and the system identifies print jobs submitted during that time frame and associates them with that task.
In one embodiment, users can provide annotations when submitting print jobs. In this case, the annotations may be integrated into the existing printing selection process, e.g., within one of the already existing notification pop-up windows informing the user that his print job has been sent to or processed by the printer. In one study, it was shown that at least a significant portion of users would have been motivated to do so to pinpoint paper-based processes that should evolve to digital form (e.g., legal documents or forms requiring a signature).
Users can also provide annotations of print jobs or time frames at a later time from a print history view. In one embodiment, a graphical user interface which provides a Personal Assessment Tool (PAT), as described above, provides a print history view visualizing the user's print jobs over time. For example, the print history provides the document title and length. In addition, users may be provided with access to the visual document content, i.e. the document page images. From this information, users can associate a set of print jobs to the task to which they belong. Alternatively, users can specify a time frame and associate it to one or a set of tasks or to a particular event generating associated tasks. This indicates that the print jobs they initiated in this time frame correspond to the tasks they were primarily executing in that time frame.
According to an exemplary embodiment, the features extracted from the print jobs, such as the visual features associated to each print job, enable them to be automatically grouped into clusters. Each cluster can be considered as corresponding to a different task or note category. This helps to detect documents involved in the same process or task, since they are often associated with documents of similar structure. For example, it may be expected that documents associated with organizing travel (plane e-tickets, hotel reservations, travel map, etc.) or with the filing of intellectual property documents (invention disclosure, patent applications, copyright forms, publications) may occur more frequently in some groups than others.
Based on features that are extracted for each document and the subset of annotated documents, the annotation component of the system learns clustering parameters for a set of clusters and propagates the labels to all the documents which have not yet been labeled. This may be performed using a supervised learning technique based on existing labels or a semi-supervised learning method.
In one exemplary embodiment, the labeled print job data can be used to identify parameters of clusters for the clustering model, which is then used to assign unlabeled print jobs to clusters based on their extracted features.
In another embodiment, the print job clustering system produces clusters of similar print jobs, initially roughly grouping, for example, print jobs related to similar basic types of documents, e.g. forms, letters, emails, presentations, etc. These initial clusters can then be refined, validated and associated to the corresponding tasks using the labels or other information input from the users who issued the jobs. Crowd sourcing information from the users, lets them annotate a small portion of their print jobs indicating to which task they correspond and also why the document was required to be in paper form. The system then uses the collected information to improve the clustering and this process can iterate until the results obtained are consistent. This approach has the advantage of requiring only a limited number of annotations and thus only a limited number of users annotating their jobs. The number of annotations needed may depend on the number of different tasks within the organization, the variability of corresponding documents involved, and on the quality of the clustering mechanism.
Once the clustering parameters are learned, unlabeled print jobs can be automatically assigned to clusters based on their print job representations alone.
In
These tablets 402, 408 are equipped with proximity sensors, such as a Bluetooth® or eBeacon® (when linked to a specific place in the room), to detect the proximity with other tablets and enable a collaborative labelling mode. With this capability, participants can move across space, and work in collaborative groups 502, 504 and share the screen to support group work and discussion.
The exemplary method employed with the hardware system performs multiple tasks. The method collects and analyses a set of printed documents to cluster them by similarity and proposes an optimal list of participants to cover the resulting clusters. It displays documents on the participants' tablets 402 by similarity to simplify their labeling suggesting labels when possible. Documents are fully readable for the owner and obfuscated for others during the collaborative labelling mode. It monitors the individual participant's 404 progression and speed to alert the consultant when a participant is blocked and suggests groups of participants that may work together when they have printed similar documents in a cluster. The method identifies possibly problematic labels given by the participants to ask for clarification and propagates knowledge captured during the labelling experience to un-labelled documents to illustrate the impact of the workshop effort on the overall set of printed documents. Lastly, it synthetizes the intermediate status and final results of the workshop at its various stages and supports live discussion between the consultant and the participants.
Before data collection begins, the system tracks and captures all the documents printed within the target organization and stores this information along with page images and meta-data (document owner) for further processing during the workshop or during its preparation. With respect to
The idea of the workshop is to collaboratively identify paper-intensive processes within a short time frame. To be able to realistically and effectively achieve this objective, the number of workshop participants 606 on one hand and the number of documents 624 the participants may be asked to label on the other hand must first be limited to realistic and reasonable values. To effectively select the workshop participants and respective documents to label, the system provides support for document clustering, identifying user contribution, and selecting clusters and participants.
In selecting the appropriate list of documents to be used, the system provides support for document clustering. During document clustering, the system clusters the whole set of documents into N clusters, identifying at the same time for each cluster a set of representative documents whose labeling allows to cover the major part of the cluster. Based on the owners of the selected documents, the system identifies for each cluster the set of contributing users with their amount of contribution (in % of documents/pages) and proposes for each cluster a selection of participants required to cover the cluster (entirely or at least a significant portion). Lastly, given the maximum number of participants for the workshop and the maximum number of documents each participant may be asked to label, the system then proposes the clusters to consider and the participants to include in the workshop. Indeed it may be impossible to aim for total coverage, and preferable instead to consider and focus only on some key clusters in a (first) workshop.
With respect to
With respect to
The system keeps track of the overall number of documents/pages removed by all the participants in this phase: they are summed up and mentioned in the final report, as X% of documents/pages printed in a non-work related context.
After the participants have agreed to participant and the proposed documents have been reviewed and determined relevant for the workshop, the document classification workshop can begin. The workshop itself consists again of several phases, the document labeling phase, see
During the workshop, and fed by the system, a large screen display permanently shows information about the workshop's actual progression and status. This display also allows the workshop facilitator to gather all the participants around it at the key moments of the workshop, and to animate the discussions that involve the whole group.
Besides this large display the facilitator also has a private display
Concerning the participants, for the duration of the workshop each of them receives a tablet, allowing him/her to interact with the system and the other participants. This tablet gives each participant access to his/her personal documents that he/she is supposed to label, i.e., to the documents he/she printed during the observation period and has pre-screened before the workshop.
Labeling phase.
The objective of the labeling phase is to label all the documents selected for the workshop with (1) the process to which they belong, e.g., billing, (2) their document type, e.g. letter, and (3) their print reason, i.e. archive, annotate, distribute, read, or sign.
With respect to
Each participant has a corresponding view of the clusters displayed on his personal tablet, augmented furthermore with a suggestion provided by the system on a cluster to start 1100, see
On this labeling screen the participant's documents are ordered by similarity, i.e., visually similar documents appear side-by-side. At the same time the biggest sets of similar documents appear grouped at the top. This facilitates the selection of large sets of documents at a time for labeling them together, all-in-one. This allows to progress significantly faster than by labeling documents in a one-by-one fashion.
Whenever the users selects one or more documents on this labeling screen, the system automatically suggests labels for those documents, in particular for the document type, an attribute that is indeed in general very much determined by and correlated with the visual appearance and similarity of documents one with the other. If similar documents have already been associated with a given document type also the corresponding process and print reason can be proposed to the participant to facilitate labeling. However, the participant can always accept or change these suggested values and enter different values as free text 1200 as shown in
Whenever this labeling process becomes tedious or cumbersome for the participant he or she can decide to stop working on the current cluster and return to the cluster visualization to select another one. At that point of time, i.e. whenever a participant leaves a cluster, the system restarts a new clustering process only with the remaining un-labeled documents, i.e., removing all documents that have in the meanwhile been labeled by all the participants. Thus, the partition in clusters evolves each time a participant leaves a cluster, and the participant will return to a new clustering view, different from the one accessed in the previous round and re-organizing the remaining documents according to their visual similarity.
With respect to
Participants can work individually as described above. However this may become tedious, especially when the size of the sets of similar documents that can be labeled together in one shot becomes too small. In that case, participants can also work together as a group, i.e., visualize and label all their documents in a common view, shared across their personal tablets. Participants may create or join an already existing group of participants at any time if they have documents that belong to a common cluster.
With further reference to
With reference to
In order to recognize a group and simplify the communication between participants to qualify documents, people need to be physically close. Several well-known technical solutions, such as eBeacon®, or direct Bluetooth®, can be used to detect tablet proximity and automatically share screens of participants close to one another. This required proximity facilitates also the oral discussion during the labeling process.
Requalification phase.
With reference to
To address these issues, the system analyses the words or text used by the participants. It uses fuzzy matching to regroup similar words in order to cover possible typos in the labels specified by the participants. It may furthermore use linguistic and/or domain specific tools to check for synonyms or expressions conveying the same or similar meaning. It also checks if document sets that are visually very similar have been labeled with different document type labels by different participants, or if the same label has been used for very different document sets. All these situations indicate potential labelling issues. Finally, the system identifies cases where similar document sets have been labeled with the same process and document type but with different print reasons: a typical confusion occurs often with respect to the archive and distribute print reasons. The system will flag all these cases so that they may be considered and discussed by the participants collectively and collaboratively in the requalification phase.
With reference to
Propagation Phase
With respect to
The system then applies the resulting classifiers to all the remaining (un-labeled) documents from the tracking period. As a result it highlights the proportion of documents that can be classified with sufficient confidence 1804, 1806, 1808. The system may use different ways to evaluate this confidence, e.g., entropy, a fixed confidence threshold etc. The proportion of documents classified with sufficient confidence directly represents the impact that the participants' labeling effort has on the global print volume, see
To motivate and reward the participants for their contribution, the facilitator introduces the propagation effect and invites participants to gather around the main screen where the propagation effect is shown through a visual animation highlighting the effect of a small set of labelled document on the global mass of documents. This illustrates the impact of the work done by all the participants.
Discussion of selected processes.
With reference to
By selecting specific points or lines in the graph, the facilitator can focus the discussion with the participants on concrete paper consumption aspects. The aim of this phase is to better understand the different processes and to identify current print reduction barriers and more complex reasons to print. All this information, captured from the people living the process, help to better understand the reality of the process and to anticipate directions for paperless alternatives 2004. See also
Some portions of the detailed description herein are presented in terms of algorithms and symbolic representations of operations on data bits performed by conventional computer components, including a central processing unit (CPU), memory storage devices for the CPU, and connected display devices. These algorithmic descriptions and representations are the means used by those skilled in the data processing arts to most effectively convey the substance of their work to others skilled in the art. An algorithm is generally perceived as a self-consistent sequence of steps leading to a desired result. The steps are those requiring physical manipulations of physical quantities. Usually, though not necessarily, these quantities take the form of electrical or magnetic signals capable of being stored, transferred, combined, compared, and otherwise manipulated. It has proven convenient at times, principally for reasons of common usage, to refer to these signals as bits, values, elements, symbols, characters, terms, numbers, or the like.
It should be understood, however, that all of these and similar terms are to be associated with the appropriate physical quantities and are merely convenient labels applied to these quantities. Unless specifically stated otherwise, as apparent from the discussion herein, it is appreciated that throughout the description, discussions utilizing terms such as “processing” or “computing” or “calculating” or “determining” or “displaying” or the like, refer to the action and processes of a computer system, or similar electronic computing device, that manipulates and transforms data represented as physical (electronic) quantities within the computer system's registers and memories into other data similarly represented as physical quantities within the computer system memories or registers or other such information storage, transmission or display devices.
The exemplary embodiment also relates to an apparatus for performing the operations discussed herein. This apparatus may be specially constructed for the required purposes, or it may comprise a general-purpose computer selectively activated or reconfigured by a computer program stored in the computer. Such a computer program may be stored in a computer readable storage medium, such as, but is not limited to, any type of disk including floppy disks, optical disks, CD-ROMs, and magnetic-optical disks, read-only memories (ROMs), random access memories (RAMs), EPROMs, EEPROMs, magnetic or optical cards, or any type of media suitable for storing electronic instructions, and each coupled to a computer system bus.
The algorithms and displays presented herein are not inherently related to any particular computer or other apparatus. Various general-purpose systems may be used with programs in accordance with the teachings herein, or it may prove convenient to construct more specialized apparatus to perform the methods described herein. The structure for a variety of these systems is apparent from the description above. In addition, the exemplary embodiment is not described with reference to any particular programming language. It will be appreciated that a variety of programming languages may be used to implement the teachings of the exemplary embodiment as described herein.
A machine-readable medium includes any mechanism for storing or transmitting information in a form readable by a machine (e.g., a computer). For instance, a machine-readable medium includes read only memory (“ROM”); random access memory (“RAM”); magnetic disk storage media; optical storage media; flash memory devices; and electrical, optical, acoustical or other form of propagated signals (e.g., carrier waves, infrared signals, digital signals, etc.), just to mention a few examples.
The methods illustrated throughout the specification, may be implemented in a computer program product that may be executed on a computer. The computer program product may comprise a non-transitory computer-readable recording medium on which a control program is recorded, such as a disk, hard drive, or the like. Common forms of non-transitory computer-readable media include, for example, floppy disks, flexible disks, hard disks, magnetic tape, or any other magnetic storage medium, CD-ROM, DVD, or any other optical medium, a RAM, a PROM, an EPROM, a FLASH-EPROM, or other memory chip or cartridge, or any other tangible medium from which a computer can read and use.
Alternatively, the method may be implemented in transitory media, such as a transmittable carrier wave in which the control program is embodied as a data signal using transmission media, such as acoustic or light waves, such as those generated during radio wave and infrared data communications, and the like.
It will be appreciated that variants of the above-disclosed and other features and functions, or alternatives thereof, may be combined into many other different systems or applications. Various presently unforeseen or unanticipated alternatives, modifications, variations or improvements therein may be subsequently made by those skilled in the art which are also intended to be encompassed by the following claims.
