1. A method comprising:
processing multiple resources to build a word dictionary configured to enable summarizing a plurality of microblogs;
using the word dictionary to create concepts, at least some individual concepts comprising a semantic tag comprising multiple words;
assigning a plurality of microblogs to a plurality of the concepts effective to form potential clusters;
computing a membership score for each microblogcluster pairing; and
using the membership score to assign a microblog to a cluster.
2. The method of claim 1, wherein said processing multiple resources comprises using a co-occurrence technique to build the word dictionary.
3. The method of claim 1, wherein said microblogs comprise company-related microblogs.
4. The method of claim 1, wherein processing the multiple resources to build a word dictionary comprises building a word dictionary that contains nouns associated with a particular domain.
5. The method of claim 1, wherein processing the multiple resources to build a word dictionary comprises:
building a word dictionary that contains nouns associated with companies;
using a pre-defined window of words to select words before and after mentions associated with a company; and
computing an association measure for at least some individual words, the association measure reflecting how often the word appears in conjunction with a company mention.
6. The method of claim 1, wherein using the word dictionary to create concepts comprises utilizing a hypernym path.
7. The method of claim 1, wherein using the word dictionary to create concepts comprises utilizing a hypernym path, the semantic tag comprising multiple words from the hypernym path.
8. The method of claim 1, wherein using the word dictionary to create concepts comprises utilizing a hypernym path, the semantic tag comprising multiple words from the hypernym path, the multiple words appearing at different hierarchical levels of the hypernym path.
9. One or more computer readable storage media comprising computer readable instructions which, when executed, implement a method comprising:
ranking individual clusters using an entropy measure on mapping words associated with the cluster, the entropy measure incorporating sentiment values associated with the mapping words, the clusters being associated with microblogs that are to be summarized; and
summarizing each cluster’s contents.
10. The one or more computer readable storage media of claim 9, wherein said summarizing comprises selecting one or more sentences based on a probability measure that incorporates a degree of sentiment relative to words that appear in the one or more sentences.
11. The one or more computer readable storage media of claim 9, wherein the microblogs are associated with one or more companies.
12. The one or more computer readable storage media of claim 9, wherein the ranking comprises computing probabilities of mapping words, wherein a mapping word’s probability incorporates an associated sentiment value.
13. The one or more computer readable storage media of claim 9, wherein the ranking comprises computing probabilities of mapping words, wherein a mapping word’s probability incorporates an associated sentiment value, sentiment values being associated with classifications of a microblog into positive, negative, or neutral categories.
14. The one or more computer readable storage media of claim 9, wherein the ranking comprises:
computing probabilities of mapping words, wherein a mapping word’s probability incorporates an associated sentiment value; and
computing entropy measures that take into account probabilities associated with the cluster’s mapping words.
15. The one or more computer readable storage media of claim 9, wherein the ranking comprises:
computing probabilities of mapping words, wherein a mapping word’s probability incorporates an associated sentiment value; and
computing entropy measures that take into account probabilities associated with the cluster’s mapping words; and
further comprising weighting computed entropy measures to reduce the entropy of at least some clusters.
16. The one or more computer readable storage media of claim 9, wherein said summarizing comprises generating a headline for a cluster, wherein the headline comprises a sorted list of mapping words that define headline words.
17. The one or more computer readable storage media of claim 9, wherein said summarizing comprises generating a headline for a cluster, wherein the headline comprises a sorted list of mapping words that define headline words; and
further comprising selecting microblogs for the cluster biased toward microblogs containing the headline words.
18. The one or more computer readable storage media of claim 9, wherein the individual microblogs have 1000 or less characters.
19. The one or more computer readable storage media of claim 9, wherein the individual microblogs have 140 characters or less.
20. A system comprising:
a learning component comprising:
a word finding component configured to process multiple resources and compute a measure of how often individual words appear in conjunction with a company name;
a word dictionary configured to contain nouns that are identified by the word finding component;
a word mapping component configured to map words appearing in the word dictionary, using a hierarchical hypernym path, to create concepts;
a concept store configured to store concepts that are created by the word mapping component, the concepts being configured to enable summarization of microblogs having 1000 characters or less; and
a summarization component comprising:
a clustering component configured to form individual clusters based on input from a plurality of microblogs for a particular company and concepts from the concept store;
a sentiment prediction component configured to compute a sentiment value for individual microblogs;
a cluster ranking component configured to rank individual clusters from the clustering component using sentiment values computed by the sentiment prediction component and probability of words computed over microblogs in a particular cluster; and
a cluster summarization component configured to provide a summary of individual clusters.
The claims below are in addition to those above.
All refrences to claim(s) which appear below refer to the numbering after this setence.
1. A mass-storage device, comprising:
a Universal Serial Bus (USB) interface;
a locking function coupled to the USB interface, wherein the locking function is accessible via a USB device class other than a mass-storage class; and
a data mass-storage memory coupled to the locking circuit.
2. The device of claim 1, wherein the data mass-storage memory comprises one of a hard disk drive, a removable storage device, an optical disk storage device, or a non-volatile semiconductor memory device.
3. The device of claim 2, wherein the locking function is implemented in a circuit.
4. The device of claim 2, wherein the locking function is implemented in firmware.
5. The device of claim 2, wherein the locking function is implemented in a universal serial bus bridge device.
6. A serial bus controller circuit, comprising:
a universal serial bus (USB) bridge controller;
a locking function coupled to the universal serial bus bridge controller; wherein the locking function is accessible via a USB device class other than a mass-storage class; and
an interface to a mass-storage device, wherein the interface is coupled to the locking function.
7. The circuit of claim 6, wherein the universal serial bus bridge controller, the locking function, and the interface to a mass storage device are located in a mass storage controller chip.
8. The circuit of claim 6, wherein the locking function is implemented in a circuit.
9. The circuit of claim 6, wherein the locking function is implemented in firmware.
10. A method for accessing a lockable mass-storage device, comprising:
connecting a lockable mass storage device to a computer;
enumerating the lockable mass storage device enumerates via an alternative interface class as a device other than a mass-storage device;
opening a \u2018keyhole\u2019 program on the computer prompting for a user to provide a password or other access means;
sending a password or other access means via the alternative interface to the lockable mass storage device; and
validating if the password or other access means is valid, and if so reNumerating the mass storage device and appearing as an unlocked device.
11. The method cited in claim 10, where an incorrect password causes reNumeration as a dummy device to prevent suspicion.
12. The method cited in claim 10 where several different passwords can be used to select between multiple virtual drives.
13. The method cited in claim 10 where an incorrect password causes a program to run on the host that can notify others of the failed password attempt.
14. A docking station apparatus, comprising:
a chassis configured for docking an electronic device;
a locking circuit coupled to the chassis, the locking circuit comprising a universal serial bus (USB) bridge function; an interface coupled to the locking circuit,
configurable to communicate between the electronic device and the locking circuit; and
a mass storage device slot coupled to the locking circuit.
15. The apparatus of claim 14, wherein the electronic device is a laptop computer.
16. The apparatus of claim 14, wherein the mass storage device slot comprises one of a hard disk drive slot, an optical disk drive slot, or a flash memory card slot.
17. The apparatus for claim 14, wherein the locking circuit is configurable to enumerating the mass storage device via an alternative interface class as a device other than a mass-storage device.
18. The apparatus of claim 17, wherein the locking circuit is configurable to request a password or other access means from the electronic device, and to validate a password or other access means received from the electronic device.
19. The apparatus of claim 18, wherein if the locking circuit validates the password or other access means, the locking circuit ReNumerates the mass storage device so it appears as an unlocked device.
20. The apparatus of claim 18, where if the locking circuit fails to validate the password or other access means, the locking circuit denies access from the electronic device to the mass storage device.
21. The apparatus of claim 18, where if the locking circuit fails to validate the password or other access means more than a threshold number of times, the locking circuit displays a dummy partition to the host device.
22. The apparatus of claim 18, wherein the locking circuit is configurable to require a different password to access each of a plurality of partitions on the mass storage device.