The patents cover many different technologies, but they mainly deal with data services like email management, online calendars and transferring web apps between devices. The patent grab, first reported by SEO by the Sea and confirmed to Mashable by Google, potentially serves two purposes: providing avenues to develop new products, and providing ammunition in litigation.
The latter reason is undoubtedly the primary one. The search giant had previously received more than 2,000 patents from IBM over the past year, a year that's seen an unprecedented amount of lawsuits in the field of patent law. Most notoriously, Apple has sued, in one form or another, many companies involved in building devices for the Android platform, including HTC, LG, and Motorola. Apple even won a recent case, which effectively forces HTC to stop selling certain devices in the U.S. or create a software fix.
By shoring up its patent portfolio, Google will be that much stronger when in its defense of Android. A big reason behind the company's $12.5 billion acquisition of Motorola Mobility this past summer was to get its hands on Motorola's large cache of patents (24,000, including pending ones).
On the product development side, it's anyone's guess which of the IBM patents may become future products, but social-media watchers might be interested in U.S. Patent 7,865,592: "Using semantic networks to develop a social network."
The patent details how a social network could be leveraged to lead users to find "experts" or like-minded enthusiasts on specific topics. As the patent's summary describes, "A method, apparatus and program product are provided for identifying common interests between users of a communication network. ... [Interests] may be determined, for example, by calculating a ratio of the number of words in a content source to the time spent viewing the content."
For example, you may want to find someone knowledgeable about real estate in a specific neighborhood to ask for buying advice. But the right person may not list that as an "interest," so you may not be able to easily find them in your expanded network (which includes friends of friends). A semantic network would find the right person to talk to by analyzing which people in your network post content having to do with the specific topic, and how much time others spend reading it.