WO2013185855A1 - A method of assigning a topic tag - Google Patents

A method of assigning a topic tag Download PDF

Info

Publication number
WO2013185855A1
WO2013185855A1 PCT/EP2012/069877 EP2012069877W WO2013185855A1 WO 2013185855 A1 WO2013185855 A1 WO 2013185855A1 EP 2012069877 W EP2012069877 W EP 2012069877W WO 2013185855 A1 WO2013185855 A1 WO 2013185855A1
Authority
WO
WIPO (PCT)
Prior art keywords
topic
tag
user
assigned
text message
Prior art date
Application number
PCT/EP2012/069877
Other languages
French (fr)
Inventor
Kareem Darwish
Ali EL-KAHKI
Original Assignee
Qatar Foundation
Hoarton, Lloyd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Qatar Foundation, Hoarton, Lloyd filed Critical Qatar Foundation
Publication of WO2013185855A1 publication Critical patent/WO2013185855A1/en

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/30—Information retrieval; Database structures therefor; File system structures therefor of unstructured textual data
    • G06F16/38—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually

Definitions

  • the present invention relates to a method of assigning a topic tag to an electronic text message, and more particularly relates to a method of assigning a topic tag to a microblog.
  • Microblogging involves the posting of electronic text messages by user which are usually in the form of short sentences, commonly referred to as microblogs.
  • a user posts a microblog via a microblogging platform to share the microblog with other users.
  • Twitter is one of the most commonly used microblogging platforms which allows users to post microblogs, known as tweets, which comprise an electronic text message consisting of up to 140 text characters.
  • a user often includes a topic tag, or hashtag, in microblogs to help other users find microblogs of interest.
  • a topic tag is usually included in a microblog by placing a hash symbol (#) in front of a word or phrase to mark the word or phrase as a topic tag or hashtag.
  • Top tags are normally defined by the users themselves and they do not adhere to a pre-defined taxonomy. This can lead to a microblog author using topic tags which do not allow microblogs to be identified easily because the topic tags are too specific, too general , unpopular, often used for spam, outdated etc.
  • Microblogs are becoming a common form of sharing information and many millions of microblogs are posted around the world each day. There is a need to improve the relevance of topic tags used in microblogs to enable information to be shared more effectively.
  • the present invention seeks to provide an improved method of assigning a topic tag to an electronic text message.
  • One aspect of the present invention provides a method of assigning a topic tag to an electronic text message, the method comprising: detecting a topic tag which is pre-assigned to the text message; analysing the text message using a topic modelling technique to identify at least one topic which is related to the content of the text message; identifying at least one topic tag which defines the at least one identified topic; comparing the pre-assigned topic tag with the topic identified using the topic modell ing technique; providing feedback information to a user regarding the comparison and an indication of whether the pre-assigned topic tag or at least one of the identified topic tags should be assigned to the electronic text message.
  • the electronic text message is a microblog.
  • the topic modelling technique is selected from a group consisting of latent semantic analysis, probabilistic latent semantic analysis, latent Dirichlet allocation and na ' fve Bayesian classification.
  • the method further comprises suggesting to the user at least one topic tag which is more relevant to the topic of the electronic text message than the pre-assigned topic tag.
  • the method further comprises suggesting to the user at least one further topic tag identified using the topic modelling technique which is related to the same topic as the pre-assigned topic tag.
  • the method further comprises comparing the pre-assigned topic tag with a plurality of topic tags available to the user via an electronic text message system and providing an indication to the user of the popularity of the pre-assigned topic tag when compared with the plurality of further topic tags available to the user via the electronic text message system.
  • the method further comprises suggesting to the user at least one further topic tag which is more popular in the electronic text message system than the user-defined topic tag.
  • the method further comprises comparing the user-defined topic tag with at least one topic tag which is categorised as electronic spam and providing an indication to the user if the user-defined topic tag at least partly matches a topic tag categorised as electronic spam.
  • Figure 1 is a schematic diagram illustrating the relationship between the latent topic of a microblog and a topic tag
  • Figure 2 is a schematic diagram showing a comparison between a good topic tag and a spam topic tag
  • Figure 3 is a schematic diagram illustrating a similar distribution of probability masses of a topic tag for two similar topic tags
  • Figure 4 is a schematic diagram illustrating a similar distribution of probability masses of a topic tag where a first topic tag is more specific than a second topic tag.
  • An embodiment of the present invention analyses an electronic text message, referred to hereinafter as a microblog.
  • a microblog incorporates raw message text and can also incorporate one or more of an explicit initial topic tag (a hashtag), an explicit ID marker (@) a timestamp or source information (@).
  • An embodiment of the invention detects an explicit initial topic tag which is pre- assigned to a microblog, for instance by the author of the microblog.
  • the raw electronic text in the microblog is then analysed using a topic modelling technique to identify latent or hidden topics related to the content of the microblog.
  • a topic modelling technique typically uses graphical models, such as Latent Dirichlet Allocation (LDA) to attempt to find latent or hidden topics in the text of a microblog.
  • LDA Latent Dirichlet Allocation
  • Other topic modelling techniques which are employed by embodiments of the invention include latent semantic analysis, probabilistic latent semantic analysis and na ' fve Bayesian classification.
  • Topic modelling is performed on microblogs automatically or in a semi- supervised manner.
  • Microblogs are assumed to be generated using at least one latent topic which, in turn, is/are produced by one or more topic tags included in each microblog.
  • Microblogs are usually very short (e.g. up to 140 characters) and so one can assume that each microblog is generated using only one latent topic. Longer text documents can, however, be generated using multiple latent topics.
  • a method of an embodiment of the invention uses a topic modelling technique to identify at least one topic which is relevant to the content of the microblog and identifies at least one topic tag which defines the at least one identified topic.
  • Embodiments of the invention use graphical models for topic modelling. The graphical models can be traversed in any direction so that a latent topic can be used to identify words or appropriate topic tags.
  • a pre- assigned topic tag can be analysed using a topic modelling technique to detect latent topics in a microblog.
  • the latent topic linking the words in a microblog and a visible topic tag is shown in figure 1 .
  • Embodiments of the invention use topic modelling techniques to detect the link between the words in a microblog and a pre- assigned topic tag.
  • Embodiments of the invention provide feedback information to a user regarding the topic tag identified using topic modelling to assist the user in tagging the microblog with at least one topic tag.
  • One embodiment of the invention compares the topic identified from a microblog using topic modelling with each pre-assigned topic tag and suggests alternative topic tags to the user.
  • the user can then tag the microblog with at least one of the identified topic tags to ensure that the microblog is tagged with at least one topic tag which is relevant to the at least one latent topic of the microblog.
  • the method therefore suggests suitable topic tags to a user to allow the topic tags to be focussed more specifically on the latent topic than a broad topic tag selected by the user or to suggest topic tags with a broader focus than topic tags selected by the user which are too specific. This allows a user to select topic tags with an appropriate relevance to maximise the chance of the topic tag being identified by other users interested in the or each latent topic of a microblog.
  • the method compares each pre-assigned topic tag with other topic tags available to users via a microblogging platform.
  • Two topic tags are equivalent if their latent topic distributions are similar.
  • a similar distribution of two topic tags is illustrated in figure 2 where the line weights reflect the relative magnitude of a topic for each tag.
  • the method can detect which of the two similar tags is more popular by counting the number of topics that are related to each tag.
  • the latent topic distribution of a pre-assigned topic tag is analysed using a topic modelling technique and compared with other topic tags available to users via the microblogging system.
  • the method identifies that the user-defined topic tag is a topic tag commonly used by spammers.
  • Figure 3 illustrates the probability mass distribution of a good topic tag and a topic tag used by spammers.
  • the line weights in figure 3 represent the relative magnitude of a topic for each tag.
  • Figure 4 illustrates the mass distributions of two topic tags. Again, the line weights in figure 4 represent the relative magnitude of a topic for each tag.
  • Topic tag 1 is more specific than topic tag 2 because the probability mass of topic tag 1 is distributed on a fewer number of topics with a higher probability for each topic than topic tag 2.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Library & Information Science (AREA)
  • Data Mining & Analysis (AREA)
  • Databases & Information Systems (AREA)
  • Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Information Retrieval, Db Structures And Fs Structures Therefor (AREA)

Abstract

A microblog is a short text message that usually comprises a topic tag that enables others to locate the microblog. A method of assigning a topic tag to a microblog comprises detecting a topic tag which is pre-assigned to the microblog. The microblog is analysed using a topic modelling technique to identify at least one topic which is related to the microblog. The method compares the pre-assigned topic tag with the topic identified using the topic modelling technique and provides feedback to a user and an indication of whether the pre-assigned topic tag or at least one of the identified topic tags should be assigned to the microblog.

Description

Title: A Method of Assigning a Topic Tag Description of Invention
The present invention relates to a method of assigning a topic tag to an electronic text message, and more particularly relates to a method of assigning a topic tag to a microblog.
Microblogging involves the posting of electronic text messages by user which are usually in the form of short sentences, commonly referred to as microblogs. A user posts a microblog via a microblogging platform to share the microblog with other users. Twitter is one of the most commonly used microblogging platforms which allows users to post microblogs, known as tweets, which comprise an electronic text message consisting of up to 140 text characters.
A user often includes a topic tag, or hashtag, in microblogs to help other users find microblogs of interest. A topic tag is usually included in a microblog by placing a hash symbol (#) in front of a word or phrase to mark the word or phrase as a topic tag or hashtag. Top tags are normally defined by the users themselves and they do not adhere to a pre-defined taxonomy. This can lead to a microblog author using topic tags which do not allow microblogs to be identified easily because the topic tags are too specific, too general , unpopular, often used for spam, outdated etc.
Microblogs are becoming a common form of sharing information and many millions of microblogs are posted around the world each day. There is a need to improve the relevance of topic tags used in microblogs to enable information to be shared more effectively. The present invention seeks to provide an improved method of assigning a topic tag to an electronic text message. One aspect of the present invention provides a method of assigning a topic tag to an electronic text message, the method comprising: detecting a topic tag which is pre-assigned to the text message; analysing the text message using a topic modelling technique to identify at least one topic which is related to the content of the text message; identifying at least one topic tag which defines the at least one identified topic; comparing the pre-assigned topic tag with the topic identified using the topic modell ing technique; providing feedback information to a user regarding the comparison and an indication of whether the pre-assigned topic tag or at least one of the identified topic tags should be assigned to the electronic text message.
Preferably, the electronic text message is a microblog.
Conveniently, the topic modelling technique is selected from a group consisting of latent semantic analysis, probabilistic latent semantic analysis, latent Dirichlet allocation and na'fve Bayesian classification.
Advantageously, the method further comprises suggesting to the user at least one topic tag which is more relevant to the topic of the electronic text message than the pre-assigned topic tag.
Preferably, the method further comprises suggesting to the user at least one further topic tag identified using the topic modelling technique which is related to the same topic as the pre-assigned topic tag. Conveniently, the method further comprises comparing the pre-assigned topic tag with a plurality of topic tags available to the user via an electronic text message system and providing an indication to the user of the popularity of the pre-assigned topic tag when compared with the plurality of further topic tags available to the user via the electronic text message system. Advantageously, the method further comprises suggesting to the user at least one further topic tag which is more popular in the electronic text message system than the user-defined topic tag.
Preferably, the method further comprises comparing the user-defined topic tag with at least one topic tag which is categorised as electronic spam and providing an indication to the user if the user-defined topic tag at least partly matches a topic tag categorised as electronic spam.
In order that the invention may be more readily understood, and so that further features thereof may be appreciated, embodiments of the invention will now be described, by way of example, with reference to the accompanying drawings in which:
Figure 1 is a schematic diagram illustrating the relationship between the latent topic of a microblog and a topic tag,
Figure 2 is a schematic diagram showing a comparison between a good topic tag and a spam topic tag, Figure 3 is a schematic diagram illustrating a similar distribution of probability masses of a topic tag for two similar topic tags, and
Figure 4 is a schematic diagram illustrating a similar distribution of probability masses of a topic tag where a first topic tag is more specific than a second topic tag. Correct microblog authoring has commercial value and so many authors, particularly commercial entities, wish to improve the visibility of their microblogs. Using correct topic tags can greatly enhance the discoverability and hence visibility of microblogs.
An embodiment of the present invention analyses an electronic text message, referred to hereinafter as a microblog. A microblog incorporates raw message text and can also incorporate one or more of an explicit initial topic tag (a hashtag), an explicit ID marker (@) a timestamp or source information (@).
An embodiment of the invention detects an explicit initial topic tag which is pre- assigned to a microblog, for instance by the author of the microblog. The raw electronic text in the microblog is then analysed using a topic modelling technique to identify latent or hidden topics related to the content of the microblog. A topic modelling technique typically uses graphical models, such as Latent Dirichlet Allocation (LDA) to attempt to find latent or hidden topics in the text of a microblog. Other topic modelling techniques which are employed by embodiments of the invention include latent semantic analysis, probabilistic latent semantic analysis and na'fve Bayesian classification.
Topic modelling is performed on microblogs automatically or in a semi- supervised manner. Microblogs are assumed to be generated using at least one latent topic which, in turn, is/are produced by one or more topic tags included in each microblog. Microblogs are usually very short (e.g. up to 140 characters) and so one can assume that each microblog is generated using only one latent topic. Longer text documents can, however, be generated using multiple latent topics.
A method of an embodiment of the invention uses a topic modelling technique to identify at least one topic which is relevant to the content of the microblog and identifies at least one topic tag which defines the at least one identified topic. Embodiments of the invention use graphical models for topic modelling. The graphical models can be traversed in any direction so that a latent topic can be used to identify words or appropriate topic tags. In addition, a pre- assigned topic tag can be analysed using a topic modelling technique to detect latent topics in a microblog.
The latent topic linking the words in a microblog and a visible topic tag is shown in figure 1 . Embodiments of the invention use topic modelling techniques to detect the link between the words in a microblog and a pre- assigned topic tag.
Embodiments of the invention provide feedback information to a user regarding the topic tag identified using topic modelling to assist the user in tagging the microblog with at least one topic tag.
One embodiment of the invention compares the topic identified from a microblog using topic modelling with each pre-assigned topic tag and suggests alternative topic tags to the user. The user can then tag the microblog with at least one of the identified topic tags to ensure that the microblog is tagged with at least one topic tag which is relevant to the at least one latent topic of the microblog. The method therefore suggests suitable topic tags to a user to allow the topic tags to be focussed more specifically on the latent topic than a broad topic tag selected by the user or to suggest topic tags with a broader focus than topic tags selected by the user which are too specific. This allows a user to select topic tags with an appropriate relevance to maximise the chance of the topic tag being identified by other users interested in the or each latent topic of a microblog.
In another embodiment, the method compares each pre-assigned topic tag with other topic tags available to users via a microblogging platform. Two topic tags are equivalent if their latent topic distributions are similar. A similar distribution of two topic tags is illustrated in figure 2 where the line weights reflect the relative magnitude of a topic for each tag. The method can detect which of the two similar tags is more popular by counting the number of topics that are related to each tag. In a further embodiment, the latent topic distribution of a pre-assigned topic tag is analysed using a topic modelling technique and compared with other topic tags available to users via the microblogging system. If the latent topic distribution of the pre-assigned tag is uniform or almost uniform or the probability mass is distributed evenly on many topics, then the method identifies that the user-defined topic tag is a topic tag commonly used by spammers. Figure 3 illustrates the probability mass distribution of a good topic tag and a topic tag used by spammers. The line weights in figure 3 represent the relative magnitude of a topic for each tag. Figure 4 illustrates the mass distributions of two topic tags. Again, the line weights in figure 4 represent the relative magnitude of a topic for each tag. Topic tag 1 is more specific than topic tag 2 because the probability mass of topic tag 1 is distributed on a fewer number of topics with a higher probability for each topic than topic tag 2.
When used in this specification and claims, the terms "comprises" and "comprising" and variations thereof mean that the specified features, steps or integers are included . The terms are not to be interpreted to exclude the presence of other features, steps or components.

Claims

Claims
1 . A method of assigning a topic tag to an electronic text message, the method comprising:
detecting a topic tag which is pre-assigned to the text message;
analysing the text message using a topic modelling technique to identify at least one topic which is related to the content of the text message;
identifying at least one topic tag which defines the at least one identified topic;
comparing the pre-assigned topic tag with the topic identified using the topic modelling technique;
providing feedback information to a user regarding the comparison and an indication of whether the pre-assigned topic tag or at least one of the identified topic tags should be assigned to the electronic text message.
2. A method according to claim 1 , wherein the electronic text message is a microblog.
3. A method according to claim 1 or claim 2, wherein the topic modelling technique is selected from a group consisting of latent semantic analysis, probabilistic latent semantic analysis, latent dirichlet allocation and na'fve Bayesian classification.
4. A method according to any one of the preceding claims, wherein the method further comprises suggesting to the user at least one topic tag which is more relevant to the topic of the electronic text message than the pre-assigned topic tag.
5. A method according to any one of the preceding claims, wherein the method further comprises suggesting to the user at least one further topic tag identified using the topic modelling technique which is related to the same topic as the pre-assigned topic tag.
6. A method according to any one of the preceding claims wherein the method further comprises comparing the pre-assigned topic tag with a plurality of topic tags available to the user via an electronic text message system and providing an indication to the user of the popularity of the pre-assigned topic tag when compared with the plurality of further topic tags available to the user via the electronic text message system.
7. A method according to claim 6, wherein the method further comprises suggesting to the user at least one further topic tag which is more popular in the electronic text message system than the user-defined topic tag.
8. A method according to any one of the preceding claims, wherein the method further comprises comparing the user-defined topic tag with at least one topic tag which is categorised as electronic spam and providing an indication to the user if the user-defined topic tag at least partly matches a topic tag categorised as electronic spam.
PCT/EP2012/069877 2012-06-15 2012-10-08 A method of assigning a topic tag WO2013185855A1 (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
GB1210660.5 2012-06-15
GBGB1210660.5A GB201210660D0 (en) 2012-06-15 2012-06-15 Interactive assignment of topic labels of microblogs

Publications (1)

Publication Number Publication Date
WO2013185855A1 true WO2013185855A1 (en) 2013-12-19

Family

ID=46640997

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/EP2012/069877 WO2013185855A1 (en) 2012-06-15 2012-10-08 A method of assigning a topic tag

Country Status (2)

Country Link
GB (1) GB201210660D0 (en)
WO (1) WO2013185855A1 (en)

Cited By (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN104991891A (en) * 2015-07-28 2015-10-21 北京大学 Short text feature extraction method
US9561098B2 (en) 2013-03-11 2017-02-07 Abbott Medical Optics Inc. Intraocular lens that matches an image surface to a retinal shape, and method of designing same
US9579192B2 (en) 2014-03-10 2017-02-28 Amo Groningen B.V. Dual-optic intraocular lens that improves overall vision where there is a local loss of retinal function
US9931200B2 (en) 2010-12-17 2018-04-03 Amo Groningen B.V. Ophthalmic devices, systems, and methods for optimizing peripheral vision
US10010407B2 (en) 2014-04-21 2018-07-03 Amo Groningen B.V. Ophthalmic devices that improve peripheral vision
CN108536868A (en) * 2018-04-24 2018-09-14 北京慧闻科技发展有限公司 The data processing method of short text data and application on social networks
US10588738B2 (en) 2016-03-11 2020-03-17 Amo Groningen B.V. Intraocular lenses that improve peripheral vision
US11096778B2 (en) 2016-04-19 2021-08-24 Amo Groningen B.V. Ophthalmic devices, system and methods that improve peripheral vision

Non-Patent Citations (5)

* Cited by examiner, † Cited by third party
Title
DAVID LANIADO ET AL: "Making Sense of Twitter", 7 November 2010, THE SEMANTIC WEB Â ISWC 2010, SPRINGER BERLIN HEIDELBERG, BERLIN, HEIDELBERG, PAGE(S) 470 - 485, ISBN: 978-3-642-17745-3, XP019159314 *
HARVEY M: "Bayesian Latent Variable Models forthe Collaborative Web", May 2011 (2011-05-01), XP002689658, Retrieved from the Internet <URL:http://132.199.143.90/morgan/papers/Thesis-Harvey.pdf> [retrieved on 20121221] *
JAFAR M AL-KOFAHI ET AL: "Fuzzy set approach for automatic tagging in evolving software", SOFTWARE MAINTENANCE (ICSM), 2010 IEEE INTERNATIONAL CONFERENCE ON, IEEE, PISCATAWAY, NJ, USA, 12 September 2010 (2010-09-12), pages 1 - 10, XP031780837, ISBN: 978-1-4244-8630-4 *
RAMAGE D ET AL: "Labeled LDA: a supervised topic model for credit attribution in multi-labeled corpora", CONFERENCE ON EMPIRICAL METHODS IN NATURAL LANGUAGE PROCESSING, vol. 1, 6 August 2009 (2009-08-06) - 7 August 2009 (2009-08-07), Singapore, pages 248 - 256, XP002689659, ISBN: 978-1-932432-59-6, Retrieved from the Internet <URL:http://dl.acm.org/citation.cfm?id=1699543> [retrieved on 20121221] *
WETZKER R: "Graph-Based Recommendation inBroad Folksonomies", 2010, XP002689660, Retrieved from the Internet <URL:http://opus.kobv.de/tuberlin/volltexte/2010/2695/pdf/wetzker_robert.pdf> [retrieved on 20121221] *

Cited By (24)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9931200B2 (en) 2010-12-17 2018-04-03 Amo Groningen B.V. Ophthalmic devices, systems, and methods for optimizing peripheral vision
US9561098B2 (en) 2013-03-11 2017-02-07 Abbott Medical Optics Inc. Intraocular lens that matches an image surface to a retinal shape, and method of designing same
US10758340B2 (en) 2013-03-11 2020-09-01 Johnson & Johnson Surgical Vision, Inc. Intraocular lens that matches an image surface to a retinal shape, and method of designing same
US10327888B2 (en) 2014-03-10 2019-06-25 Amo Groningen B.V. Enhanced toric lens that improves overall vision where there is a local loss of retinal function
US9579192B2 (en) 2014-03-10 2017-02-28 Amo Groningen B.V. Dual-optic intraocular lens that improves overall vision where there is a local loss of retinal function
US11517423B2 (en) 2014-03-10 2022-12-06 Amo Groningen B.V. Piggyback intraocular lens that improves overall vision where there is a local loss of retinal function
US9636215B2 (en) 2014-03-10 2017-05-02 Amo Groningen B.V. Enhanced toric lens that improves overall vision where there is a local loss of retinal function
US11331181B2 (en) 2014-03-10 2022-05-17 Amo Groningen B.V. Fresnel piggyback intraocular lens that improves overall vision where there is a local loss of retinal function
US10016270B2 (en) 2014-03-10 2018-07-10 Amo Groningen B.V. Dual-optic intraocular lens that improves overall vision where there is a local loss of retinal function
US9867693B2 (en) 2014-03-10 2018-01-16 Amo Groningen B.V. Intraocular lens that improves overall vision where there is a local loss of retinal function
US10136990B2 (en) 2014-03-10 2018-11-27 Amo Groningen B.V. Piggyback intraocular lens that improves overall vision where there is a local loss of retinal function
US10143548B2 (en) 2014-03-10 2018-12-04 Amo Groningen B.V. Fresnel piggyback intraocular lens that improves overall vision where there is a local loss of retinal function
US11534291B2 (en) 2014-03-10 2022-12-27 Amo Groningen B.V. Intraocular lens that improves overall vision where there is a local loss of retinal function
US10456242B2 (en) 2014-03-10 2019-10-29 Amo Groningen B.V. Intraocular lens that improves overall vision where there is a local loss of retinal function
US10588739B2 (en) 2014-04-21 2020-03-17 Amo Groningen B.V. Ophthalmic devices, system and methods that improve peripheral vision
US11660183B2 (en) 2014-04-21 2023-05-30 Amo Groningen B.V. Ophthalmic devices, system and methods that improve peripheral vision
US10010407B2 (en) 2014-04-21 2018-07-03 Amo Groningen B.V. Ophthalmic devices that improve peripheral vision
CN104991891A (en) * 2015-07-28 2015-10-21 北京大学 Short text feature extraction method
CN104991891B (en) * 2015-07-28 2018-03-30 北京大学 A kind of short text feature extracting method
US10588738B2 (en) 2016-03-11 2020-03-17 Amo Groningen B.V. Intraocular lenses that improve peripheral vision
US11160651B2 (en) 2016-03-11 2021-11-02 Amo Groningen B.V. Intraocular lenses that improve peripheral vision
US11096778B2 (en) 2016-04-19 2021-08-24 Amo Groningen B.V. Ophthalmic devices, system and methods that improve peripheral vision
CN108536868A (en) * 2018-04-24 2018-09-14 北京慧闻科技发展有限公司 The data processing method of short text data and application on social networks
CN108536868B (en) * 2018-04-24 2022-04-15 北京慧闻科技(集团)有限公司 Data processing method and device for short text data on social network

Also Published As

Publication number Publication date
GB201210660D0 (en) 2012-08-01

Similar Documents

Publication Publication Date Title
WO2013185855A1 (en) A method of assigning a topic tag
Castillo Big crisis data: social media in disasters and time-critical situations
Yang et al. Automatic detection of rumor on sina weibo
Phuvipadawat et al. Breaking news detection and tracking in Twitter
Li et al. Question identification on twitter
Rubin et al. Deception detection for news: three types of fakes
Varshney et al. A unified approach for detection of Clickbait videos on YouTube using cognitive evidences
Becker Identification and characterization of events in social media
WO2012174637A1 (en) System and method for matching comment data to text data
CN104978332B (en) User-generated content label data generation method, device and correlation technique and device
Tolosi et al. An analysis of event-agnostic features for rumour classification in twitter
WO2012142652A1 (en) Method for identifying potential defects in a block of text using socially contributed pattern/message rules
Gayberi et al. Popularity prediction of posts in social networks based on user, post and image features
VanDam et al. Understanding compromised accounts on twitter
WO2014000130A1 (en) Method or system for automated extraction of hyper-local events from one or more web pages
Rakesh et al. Location-specific tweet detection and topic summarization in twitter
Choudhury et al. Personal life event detection from social media
Kunneman et al. Event detection in Twitter: A machine-learning approach based on term pivoting
Gorrell et al. Using@ Twitter conventions to improve# LOD-based named entity disambiguation
Chen et al. Scam detection in Twitter
Kinsella et al. Improving categorisation in social media using hyperlinks to structured data sources
Sams et al. The presence of hyperlinks on social network sites: A case study of Cyworld in Korea
Luo et al. Structuring T weets for improving T witter search
Magdy et al. Bridging social media via distant supervision
Lakshmi et al. Sentiment analysis of twitter data

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 12794883

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 12794883

Country of ref document: EP

Kind code of ref document: A1